Chang Wen Chen

Bruce X. B. Yu, Zhi Zhang, Yongxu Liu et al.

3D human pose estimation has been researched for decades with promising fruits. 3D human pose lifting is one of the promising research directions toward the task where both estimated pose and ground truth pose data are used for training. Existing pose lifting works mainly focus on improving the performance of estimated pose, but they usually underperform when testing on the ground truth pose data. We observe that the performance of the estimated pose can be easily improved by preparing good quality 2D pose, such as fine-tuning the 2D pose or using advanced 2D pose detectors. As such, we concentrate on improving the 3D human pose lifting via ground truth data for the future improvement of more quality estimated pose data. Towards this goal, a simple yet effective model called Global-local Adaptive Graph Convolutional Network (GLA-GCN) is proposed in this work. Our GLA-GCN globally models the spatiotemporal structure via a graph representation and backtraces local joint features for 3D human pose estimation via individually connected layers. To validate our model design, we conduct extensive experiments on three benchmark datasets: Human3.6M, HumanEva-I, and MPI-INF-3DHP. Experimental results show that our GLA-GCN implemented with ground truth 2D poses significantly outperforms state-of-the-art methods (e.g., up to around 3%, 17%, and 14% error reductions on Human3.6M, HumanEva-I, and MPI-INF-3DHP, respectively). GitHub: https://github.com/bruceyo/GLA-GCN.

Ye Liu, Siyuan Li, Yang Wu et al.

Finding relevant moments and highlights in videos according to natural language queries is a natural and highly valuable common need in the current video content explosion era. Nevertheless, jointly conducting moment retrieval and highlight detection is an emerging research topic, even though its component problems and some related tasks have already been studied for a while. In this paper, we present the first unified framework, named Unified Multi-modal Transformers (UMT), capable of realizing such joint optimization while can also be easily degenerated for solving individual problems. As far as we are aware, this is the first scheme to integrate multi-modal (visual-audio) learning for either joint optimization or the individual moment retrieval task, and tackles moment retrieval as a keypoint detection problem using a novel query generator and query decoder. Extensive comparisons with existing methods and ablation studies on QVHighlights, Charades-STA, YouTube Highlights, and TVSum datasets demonstrate the effectiveness, superiority, and flexibility of the proposed method under various settings. Source code and pre-trained models are available at https://github.com/TencentARC/UMT.

Bruce X. B. Yu, Jianlong Chang, Lingbo Liu et al.

Parameter efficient transfer learning (PETL) aims at making good use of the representation knowledge in the pre-trained large models by fine-tuning a small number of parameters. Recently, taking inspiration from the natural language processing (NLP) domain, popular PETL techniques such as prompt-tuning and Adapter have also been successfully applied to the vision domain. However, prefix-tuning remains under-explored for vision tasks. In this work, we intend to adapt large vision models (LVMs) to downstream tasks with a good parameter-accuracy trade-off. Towards this goal, we propose a framework with a unified view of PETL called visual-PETL (V-PETL) to investigate the effects of different PETL techniques, data scales of downstream domains, positions of trainable parameters, and other aspects affecting the trade-off. Specifically, we analyze the positional importance of trainable parameters and differences between NLP and vision tasks in terms of data structures and pre-training mechanisms while implementing various PETL techniques, especially for the under-explored prefix-tuning technique. Based on a comprehensive understanding of the differences between NLP and vision data, we propose a new variation of the prefix-tuning module called parallel attention (PATT) for vision downstream tasks. An extensive empirical analysis on vision tasks via different frozen LVMs has been carried and the findings show that the proposed PATT can effectively contribute to other PETL techniques. An effective scheme Swin-BAPAT derived from the proposed V-PETL framework achieves significantly better performance than the state-of-the-art AdaptFormer-Swin with slightly more parameters and outperforms full-tuning with far fewer parameters. Code and data are available at: https://github.com/bruceyo/V-PETL.

Tiancheng Lin, Zhimiao Yu, Hongyu Hu et al.

Multi-instance learning (MIL) is an effective paradigm for whole-slide pathological images (WSIs) classification to handle the gigapixel resolution and slide-level label. Prevailing MIL methods primarily focus on improving the feature extractor and aggregator. However, one deficiency of these methods is that the bag contextual prior may trick the model into capturing spurious correlations between bags and labels. This deficiency is a confounder that limits the performance of existing MIL methods. In this paper, we propose a novel scheme, Interventional Bag Multi-Instance Learning (IBMIL), to achieve deconfounded bag-level prediction. Unlike traditional likelihood-based strategies, the proposed scheme is based on the backdoor adjustment to achieve the interventional training, thus is capable of suppressing the bias caused by the bag contextual prior. Note that the principle of IBMIL is orthogonal to existing bag MIL methods. Therefore, IBMIL is able to bring consistent performance boosting to existing schemes, achieving new state-of-the-art performance. Code is available at https://github.com/HHHedo/IBMIL.

Yanshuo Wang, Yuan Xu, Xuesong Li et al.

Egocentric assistants often rely on first-person view data to capture user behavior and context for personalized services. Since different users exhibit distinct habits, preferences, and routines, such personalization is essential for truly effective assistance. However, effectively integrating long-term user data for personalization remains a key challenge. To address this, we introduce EgoSelf, a system that includes a graph-based interaction memory constructed from past observations and a dedicated learning task for personalization. The memory captures temporal and semantic relationships among interaction events and entities, from which user-specific profiles are derived. The personalized learning task is formulated as a prediction problem where the model predicts possible future interactions from individual user's historical behavior recorded in the graph. Extensive experiments demonstrate the effectiveness of EgoSelf as a personalized egocentric assistant. Code is available at \href{https://abie-e.github.io/egoself_project/}{https://abie-e.github.io/egoself\_project/}.

Zuyao Chen, Jinlin Wu, Zhen Lei et al.

Scene Graph Generation (SGG) offers a structured representation critical in many computer vision applications. Traditional SGG approaches, however, are limited by a closed-set assumption, restricting their ability to recognize only predefined object and relation categories. To overcome this, we categorize SGG scenarios into four distinct settings based on the node and edge: Closed-set SGG, Open Vocabulary (object) Detection-based SGG (OvD-SGG), Open Vocabulary Relation-based SGG (OvR-SGG), and Open Vocabulary Detection + Relationbased SGG (OvD+R-SGG). While object-centric open vocabulary SGG has been studied recently, the more challenging problem of relation-involved open-vocabulary SGG remains relatively unexplored. To fill this gap, we propose a unified framework named OvSGTR towards fully open vocabulary SGG from a holistic view. The proposed framework is an end-to-end transformer architecture, which learns a visual-concept alignment for both nodes and edges, enabling the model to recognize unseen categories. For the more challenging settings of relation-involved open vocabulary SGG, the proposed approach integrates relation-aware pretraining utilizing image-caption data and retains visual-concept alignment through knowledge distillation. Comprehensive experimental results on the Visual Genome benchmark demonstrate the effectiveness and superiority of the proposed framework. Our code is available at https://github.com/gpt4vision/OvSGTR/.

Xiongkuo Min, Yixuan Gao, Yuqin Cao et al.

Traditional in the wild image quality assessment (IQA) models are generally trained with the quality labels of mean opinion score (MOS), while missing the rich subjective quality information contained in the quality ratings, for example, the standard deviation of opinion scores (SOS) or even distribution of opinion scores (DOS). In this paper, we propose a novel IQA method named RichIQA to explore the rich subjective rating information beyond MOS to predict image quality in the wild. RichIQA is characterized by two key novel designs: (1) a three-stage image quality prediction network which exploits the powerful feature representation capability of the Convolutional vision Transformer (CvT) and mimics the short-term and long-term memory mechanisms of human brain; (2) a multi-label training strategy in which rich subjective quality information like MOS, SOS and DOS are concurrently used to train the quality prediction network. Powered by these two novel designs, RichIQA is able to predict the image quality in terms of a distribution, from which the mean image quality can be subsequently obtained. Extensive experimental results verify that the three-stage network is tailored to predict rich quality information, while the multi-label training strategy can fully exploit the potentials within subjective quality rating and enhance the prediction performance and generalizability of the network. RichIQA outperforms state-of-the-art competitors on multiple large-scale in the wild IQA databases with rich subjective rating labels. The code of RichIQA will be made publicly available on GitHub.

Yiyang Jiang, Wengyu Zhang, Xulu Zhang et al.

In this paper, we investigate the feasibility of leveraging large language models (LLMs) for integrating general knowledge and incorporating pseudo-events as priors for temporal content distribution in video moment retrieval (VMR) models. The motivation behind this study arises from the limitations of using LLMs as decoders for generating discrete textual descriptions, which hinders their direct application to continuous outputs like salience scores and inter-frame embeddings that capture inter-frame relations. To overcome these limitations, we propose utilizing LLM encoders instead of decoders. Through a feasibility study, we demonstrate that LLM encoders effectively refine inter-concept relations in multimodal embeddings, even without being trained on textual embeddings. We also show that the refinement capability of LLM encoders can be transferred to other embeddings, such as BLIP and T5, as long as these embeddings exhibit similar inter-concept similarity patterns to CLIP embeddings. We present a general framework for integrating LLM encoders into existing VMR architectures, specifically within the fusion module. Through experimental validation, we demonstrate the effectiveness of our proposed methods by achieving state-of-the-art performance in VMR. The source code can be accessed at https://github.com/fletcherjiang/LLMEPET.

Peipei Zhu, Xiao Wang, Lin Zhu et al.

Unpaired Image Captioning (UIC) has been developed to learn image descriptions from unaligned vision-language sample pairs. Existing works usually tackle this task using adversarial learning and visual concept reward based on reinforcement learning. However, these existing works were only able to learn limited cross-domain information in vision and language domains, which restrains the captioning performance of UIC. Inspired by the success of Vision-Language Pre-Trained Models (VL-PTMs) in this research, we attempt to infer the cross-domain cue information about a given image from the large VL-PTMs for the UIC task. This research is also motivated by recent successes of prompt learning in many downstream multi-modal tasks, including image-text retrieval and vision question answering. In this work, a semantic prompt is introduced and aggregated with visual features for more accurate caption prediction under the adversarial learning framework. In addition, a metric prompt is designed to select high-quality pseudo image-caption samples obtained from the basic captioning model and refine the model in an iterative manner. Extensive experiments on the COCO and Flickr30K datasets validate the promising captioning ability of the proposed model. We expect that the proposed prompt-based UIC model will stimulate a new line of research for the VL-PTMs based captioning.

Huairui Wang, Zhenzhong Chen, Chang Wen Chen

Learned video compression has recently emerged as an essential research topic in developing advanced video compression technologies, where motion compensation is considered one of the most challenging issues. In this paper, we propose a learned video compression framework via heterogeneous deformable compensation strategy (HDCVC) to tackle the problems of unstable compression performance caused by single-size deformable kernels in downsampled feature domain. More specifically, instead of utilizing optical flow warping or single-size-kernel deformable alignment, the proposed algorithm extracts features from the two adjacent frames to estimate content-adaptive heterogeneous deformable (HetDeform) kernel offsets. Then we transform the reference features with the HetDeform convolution to accomplish motion compensation. Moreover, we design a Spatial-Neighborhood-Conditioned Divisive Normalization (SNCDN) to achieve more effective data Gaussianization combined with the Generalized Divisive Normalization. Furthermore, we propose a multi-frame enhanced reconstruction module for exploiting context and temporal information for final quality enhancement. Experimental results indicate that HDCVC achieves superior performance than the recent state-of-the-art learned video compression approaches.

Ye Liu, Zongyang Ma, Zhongang Qi et al.

Recent advances in Video Large Language Models (Video-LLMs) have demonstrated their great potential in general-purpose video understanding. To verify the significance of these models, a number of benchmarks have been proposed to diagnose their capabilities in different scenarios. However, existing benchmarks merely evaluate models through video-level question-answering, lacking fine-grained event-level assessment and task diversity. To fill this gap, we introduce E.T. Bench (Event-Level & Time-Sensitive Video Understanding Benchmark), a large-scale and high-quality benchmark for open-ended event-level video understanding. Categorized within a 3-level task taxonomy, E.T. Bench encompasses 7.3K samples under 12 tasks with 7K videos (251.4h total length) under 8 domains, providing comprehensive evaluations. We extensively evaluated 8 Image-LLMs and 12 Video-LLMs on our benchmark, and the results reveal that state-of-the-art models for coarse-level (video-level) understanding struggle to solve our fine-grained tasks, e.g., grounding event-of-interests within videos, largely due to the short video context length, improper time representations, and lack of multi-event training data. Focusing on these issues, we further propose a strong baseline model, E.T. Chat, together with an instruction-tuning dataset E.T. Instruct 164K tailored for fine-grained event-level understanding. Our simple but effective solution demonstrates superior performance in multiple scenarios.

Tengpeng Li, Hanli Wang, Bin He et al.

As a technically challenging topic, visual storytelling aims at generating an imaginary and coherent story with narrative multi-sentences from a group of relevant images. Existing methods often generate direct and rigid descriptions of apparent image-based contents, because they are not capable of exploring implicit information beyond images. Hence, these schemes could not capture consistent dependencies from holistic representation, impairing the generation of reasonable and fluent story. To address these problems, a novel knowledge-enriched attention network with group-wise semantic model is proposed. Three main novel components are designed and supported by substantial experiments to reveal practical advantages. First, a knowledge-enriched attention network is designed to extract implicit concepts from external knowledge system, and these concepts are followed by a cascade cross-modal attention mechanism to characterize imaginative and concrete representations. Second, a group-wise semantic module with second-order pooling is developed to explore the globally consistent guidance. Third, a unified one-stage story generation model with encoder-decoder structure is proposed to simultaneously train and infer the knowledge-enriched attention network, group-wise semantic module and multi-modal story generation decoder in an end-to-end fashion. Substantial experiments on the popular Visual Storytelling dataset with both objective and subjective evaluation metrics demonstrate the superior performance of the proposed scheme as compared with other state-of-the-art methods.

Qinyu Li, Tengpeng Li, Hanli Wang et al.

Translating visual data into natural language is essential for machines to understand the world and interact with humans. In this work, a comprehensive study is conducted on video paragraph captioning, with the goal to generate paragraph-level descriptions for a given video. However, current researches mainly focus on detecting objective facts, ignoring the needs to establish the logical associations between sentences and to discover more accurate emotions related to video contents. Such a problem impairs fluent and abundant expressions of predicted captions, which are far below human language tandards. To solve this problem, we propose to construct a large-scale emotion and logic driven multilingual dataset for this task. This dataset is named EMVPC (standing for "Emotional Video Paragraph Captioning") and contains 53 widely-used emotions in daily life, 376 common scenes corresponding to these emotions, 10,291 high-quality videos and 20,582 elaborated paragraph captions with English and Chinese versions. Relevant emotion categories, scene labels, emotion word labels and logic word labels are also provided in this new dataset. The proposed EMVPC dataset intends to provide full-fledged video paragraph captioning in terms of rich emotions, coherent logic and elaborate expressions, which can also benefit other tasks in vision-language fields. Furthermore, a comprehensive study is conducted through experiments on existing benchmark video paragraph captioning datasets and the proposed EMVPC. The stateof-the-art schemes from different visual captioning tasks are compared in terms of 15 popular metrics, and their detailed objective as well as subjective results are summarized. Finally, remaining problems and future directions of video paragraph captioning are also discussed. The unique perspective of this work is expected to boost further development in video paragraph captioning research.

Peipei Zhu, Xiao Wang, Yong Luo et al.

The goal of unpaired image captioning (UIC) is to describe images without using image-caption pairs in the training phase. Although challenging, we except the task can be accomplished by leveraging a training set of images aligned with visual concepts. Most existing studies use off-the-shelf algorithms to obtain the visual concepts because the Bounding Box (BBox) labels or relationship-triplet labels used for the training are expensive to acquire. In order to resolve the problem in expensive annotations, we propose a novel approach to achieve cost-effective UIC. Specifically, we adopt image-level labels for the optimization of the UIC model in a weakly-supervised manner. For each image, we assume that only the image-level labels are available without specific locations and numbers. The image-level labels are utilized to train a weakly-supervised object recognition model to extract object information (e.g., instance) in an image, and the extracted instances are adopted to infer the relationships among different objects based on an enhanced graph neural network (GNN). The proposed approach achieves comparable or even better performance compared with previous methods without the expensive cost of annotations. Furthermore, we design an unrecognized object (UnO) loss combined with a visual concept reward to improve the alignment of the inferred object and relationship information with the images. It can effectively alleviate the issue encountered by existing UIC models about generating sentences with nonexistent objects. To the best of our knowledge, this is the first attempt to solve the problem of Weakly-Supervised visual concept recognition for UIC (WS-UIC) based only on image-level labels. Extensive experiments have been carried out to demonstrate that the proposed WS-UIC model achieves inspiring results on the COCO dataset while significantly reducing the cost of labeling.

Ruixiang Jiang, Lingbo Liu, Changwen Chen

We show that the representation of one modality can effectively guide the prompting of another modality for parameter-efficient multimodal fusion. Specifically, we first encode one modality and use its representation as a prior to conditionally prompt all frozen layers of the other modality. This is achieved by disentangling the vanilla prompt vectors into three types of specialized prompts that adaptively capture global-level and instance-level features. To better produce the instance-wise prompt, we introduce the mixture of prompt experts (MoPE) to dynamically route each instance to the most suitable prompt experts for encoding. We further study a regularization term to avoid degenerated prompt expert routing. Thanks to our design, our method can effectively transfer the pretrained knowledge in unimodal encoders for downstream multimodal tasks. Compared with vanilla prompting, we show that our MoPE-based conditional prompting is more expressive, thereby scales better with training data and the total number of prompts. We also demonstrate that our prompt tuning is architecture-agnostic, thereby offering high modularity. Extensive experiments over three multimodal datasets demonstrate state-of-the-art results, matching or surpassing the performance achieved through fine-tuning, while only necessitating 0.7% of the trainable parameters. Code will be released: https://github.com/songrise/ConditionalPrompt.

Ye Liu, Jixuan He, Wanhua Li et al.

Video temporal grounding (VTG) is a fine-grained video understanding problem that aims to ground relevant clips in untrimmed videos given natural language queries. Most existing VTG models are built upon frame-wise final-layer CLIP features, aided by additional temporal backbones (e.g., SlowFast) with sophisticated temporal reasoning mechanisms. In this work, we claim that CLIP itself already shows great potential for fine-grained spatial-temporal modeling, as each layer offers distinct yet useful information under different granularity levels. Motivated by this, we propose Reversed Recurrent Tuning ($R^2$-Tuning), a parameter- and memory-efficient transfer learning framework for video temporal grounding. Our method learns a lightweight $R^2$ Block containing only 1.5% of the total parameters to perform progressive spatial-temporal modeling. Starting from the last layer of CLIP, $R^2$ Block recurrently aggregates spatial features from earlier layers, then refines temporal correlation conditioning on the given query, resulting in a coarse-to-fine scheme. $R^2$-Tuning achieves state-of-the-art performance across three VTG tasks (i.e., moment retrieval, highlight detection, and video summarization) on six public benchmarks (i.e., QVHighlights, Charades-STA, Ego4D-NLQ, TACoS, YouTube Highlights, and TVSum) even without the additional backbone, demonstrating the significance and effectiveness of the proposed scheme. Our code is available at https://github.com/yeliudev/R2-Tuning.

Tao Yang, Haokui Zhang, Wenze Hu et al.

Transformer models have made tremendous progress in various fields in recent years. In the field of computer vision, vision transformers (ViTs) also become strong alternatives to convolutional neural networks (ConvNets), yet they have not been able to replace ConvNets since both have their own merits. For instance, ViTs are good at extracting global features with attention mechanisms while ConvNets are more efficient in modeling local relationships due to their strong inductive bias. A natural idea that arises is to combine the strengths of both ConvNets and ViTs to design new structures. In this paper, we propose a new basic neural network operator named position-aware circular convolution (ParC) and its accelerated version Fast-ParC. The ParC operator can capture global features by using a global kernel and circular convolution while keeping location sensitiveness by employing position embeddings. Our Fast-ParC further reduces the O(n2) time complexity of ParC to O(n log n) using Fast Fourier Transform. This acceleration makes it possible to use global convolution in the early stages of models with large feature maps, yet still maintains the overall computational cost comparable with using 3x3 or 7x7 kernels. The proposed operation can be used in a plug-and-play manner to 1) convert ViTs to pure-ConvNet architecture to enjoy wider hardware support and achieve higher inference speed; 2) replacing traditional convolutions in the deep stage of ConvNets to improve accuracy by enlarging the effective receptive field. Experiment results show that our ParC op can effectively enlarge the receptive field of traditional ConvNets, and adopting the proposed op benefits both ViTs and ConvNet models on all three popular vision tasks, image classification, object

Hongchen Wei, Zhihong Tan, Yaosi Hu et al.

Large Multimodal Models (LMMs) have demonstrated exceptional performance in video captioning tasks, particularly for short videos. However, as the length of the video increases, generating long, detailed captions becomes a significant challenge. In this paper, we investigate the limitations of LMMs in generating long captions for long videos. Our analysis reveals that open-source LMMs struggle to consistently produce outputs exceeding 300 words, leading to incomplete or overly concise descriptions of the visual content. This limitation hinders the ability of LMMs to provide comprehensive and detailed captions for long videos, ultimately missing important visual information. Through controlled experiments, we find that the scarcity of paired examples with long-captions during training is the primary factor limiting the model's output length. However, manually annotating long-caption examples for long-form videos is time-consuming and expensive. To overcome the annotation bottleneck, we propose the LongCaption-Agent, a framework that synthesizes long caption data by hierarchical semantic aggregation. % aggregating multi-level descriptions. Using LongCaption-Agent, we curated a new long-caption dataset, LongCaption-10K. We also develop LongCaption-Bench, a benchmark designed to comprehensively evaluate the quality of long captions generated by LMMs. By incorporating LongCaption-10K into training, we enable LMMs to generate captions exceeding 1,000 words for long-form videos, while maintaining high output quality. In LongCaption-Bench, our model achieved State-of-The-Art performance, even surpassing larger proprietary models like GPT4o.

Ruixiang Jiang, Changwen Chen

The rapid technical progress of generative art (GenArt) has democratized the creation of visually appealing imagery. However, achieving genuine artistic impact - the kind that resonates with viewers on a deeper, more meaningful level - remains formidable as it requires a sophisticated aesthetic sensibility. This sensibility involves a multifaceted cognitive process extending beyond mere visual appeal, which is often overlooked by current computational methods. This paper pioneers an approach to capture this complex process by investigating how the reasoning capabilities of Multimodal LLMs (MLLMs) can be effectively elicited to perform aesthetic judgment. Our analysis reveals a critical challenge: MLLMs exhibit a tendency towards hallucinations during aesthetic reasoning, characterized by subjective opinions and unsubstantiated artistic interpretations. We further demonstrate that these hallucinations can be suppressed by employing an evidence-based and objective reasoning process, as substantiated by our proposed baseline, ArtCoT. MLLMs prompted by this principle produce multifaceted, in-depth aesthetic reasoning that aligns significantly better with human judgment. These findings have direct applications in areas such as AI art tutoring and as reward models for image generation. Ultimately, we hope this work paves the way for AI systems that can truly understand, appreciate, and contribute to art that aligns with human aesthetic values. Project homepage: https://github.com/songrise/MLLM4Art.

Ruixiang Jiang, Changwen Chen

Artistic styles are defined by both their structural and appearance elements. Existing neural stylization techniques primarily focus on transferring appearance-level features such as color and texture, often neglecting the equally crucial aspect of structural stylization. To address this gap, we introduce \textbf{DiffArtist}, the first 2D stylization method to offer fine-grained, simultaneous control over both structure and appearance style strength. This dual controllability is achieved by representing structure and appearance generation as separate diffusion processes, necessitating no further tuning or additional adapters. To properly evaluate this new capability of dual stylization, we further propose a Multimodal LLM-based stylization evaluator that aligns significantly better with human preferences than existing metrics. Extensive analysis shows that DiffArtist achieves superior style fidelity and dual-controllability compared to state-of-the-art methods. Its text-driven, training-free design and unprecedented dual controllability make it a powerful and interactive tool for various creative applications. Project homepage: https://diffusionartist.github.io.

Tao Yang, Yingmin Luo, Zhongang Qi et al.

Layout generation is the keystone in achieving automated graphic design, requiring arranging the position and size of various multi-modal design elements in a visually pleasing and constraint-following manner. Previous approaches are either inefficient for large-scale applications or lack flexibility for varying design requirements. Our research introduces a unified framework for automated graphic layout generation, leveraging the multi-modal large language model (MLLM) to accommodate diverse design tasks. In contrast, our data-driven method employs structured text (JSON format) and visual instruction tuning to generate layouts under specific visual and textual constraints, including user-defined natural language specifications. We conducted extensive experiments and achieved state-of-the-art (SOTA) performance on public multi-modal layout generation benchmarks, demonstrating the effectiveness of our method. Moreover, recognizing existing datasets' limitations in capturing the complexity of real-world graphic designs, we propose two new datasets for much more challenging tasks (user-constrained generation and complicated poster), further validating our model's utility in real-life settings. Marking by its superior accessibility and adaptability, this approach further automates large-scale graphic design tasks. Finally, we develop an automated text-to-poster system that generates editable SVG posters based on users' design intentions, bridging the gap between layout generation and real-world graphic design applications. This system integrates our proposed layout generation method as the core component, demonstrating its effectiveness in practical scenarios. The code and datasets are open-sourced on https://github.com/posterllava/PosterLLaVA.

Tiancheng Lin, Jinglei Zhang, Yi Xu et al.

Context-aware methods have achieved remarkable advancements in supervised scene text recognition by leveraging semantic priors from words. Considering the heterogeneity of text and background in STR, we propose that such contextual priors can be reinterpreted as the relations between textual elements, serving as effective self-supervised labels for representation learning. However, textual relations are restricted to the finite size of the dataset due to lexical dependencies, which causes over-fitting problem, thus compromising the representation quality. To address this, our work introduces a unified framework of Relational Contrastive Learning and Masked Image Modeling for STR (RCMSTR), which explicitly models the enriched textual relations. For the RCL branch, we first introduce the relational rearrangement module to cultivate new relations on the fly. Based on this, we further conduct relational contrastive learning to model the intra- and inter-hierarchical relations for frames, sub-words and words. On the other hand, MIM can naturally boost the context information via masking, where we find that the block masking strategy is more effective for STR. For the effective integration of RCL and MIM, we also introduce a novel decoupling design aimed at mitigating the impact of masked images on contrastive learning. Additionally, to enhance the compatibility of MIM with CNNs, we propose the adoption of sparse convolutions and directly sharing the weights with dense convolutions in training. The proposed RCMSTR demonstrates superior performance in various evaluation protocols for different STR-related downstream tasks, outperforming the existing state-of-the-art self-supervised STR techniques. Ablation studies and qualitative experimental results further validate the effectiveness of our method. The code and pre-trained models will be available at https://github.com/ThunderVVV/RCMSTR .

Ruixiang Jiang, Lingbo Liu, Changwen Chen

Despite the demonstrated parameter efficiency of prompt-based fusion, its limited adaptivity and expressiveness hinder its effectiveness for multimodal applications at scale. In this paper, we present the first comprehensive study addressing these limitations. Our key motivation is to ``divide and conquer'' the vanilla prompt, traditionally shared across all instances, by generating instance-specific prompts. Specifically, we propose the Mixture of Prompt Experts (MoPE), a framework that significantly enhances prompt adaptivity and expressiveness by dynamically generating instance-specific prompts. MoPE leverages multimodal pairings as additional evidence, allowing the model to adaptively select optimal prompts tailored to each individual instance. Unlike traditional prompt-fusion methods, which encounter scalability bottlenecks when optimizing long unified prompts, MoPE maintains fixed prompt length while effectively scaling the number of specialized experts. Moreover, we investigate regularization terms to encourage expert specialization, resulting in highly adaptive and interpretable prompting. MoPE fundamentally changes the scaling dynamic, unlocking greater expressiveness and adaptability to complex multimodal relationships, enabling the model to selectively attend to task-relevant sub-sequences based on instance-specific multimodal input. Extensive experiments across six multimodal datasets spanning four modalities demonstrate state-of-the-art performance for multimodal fusion, matching or surpassing the performance of fine-tuning while requiring only 0.8% of the trainable parameters. Code is available: https://github.com/songrise/MoPE.

Ruixiang Jiang, Lingbo Liu, Changwen Chen

Recent advances in visual-language models have shown remarkable zero-shot text-image matching ability that is transferable to downstream tasks such as object detection and segmentation. Adapting these models for object counting, however, remains a formidable challenge. In this study, we first investigate transferring vision-language models (VLMs) for class-agnostic object counting. Specifically, we propose CLIP-Count, the first end-to-end pipeline that estimates density maps for open-vocabulary objects with text guidance in a zero-shot manner. To align the text embedding with dense visual features, we introduce a patch-text contrastive loss that guides the model to learn informative patch-level visual representations for dense prediction. Moreover, we design a hierarchical patch-text interaction module to propagate semantic information across different resolution levels of visual features. Benefiting from the full exploitation of the rich image-text alignment knowledge of pretrained VLMs, our method effectively generates high-quality density maps for objects-of-interest. Extensive experiments on FSC-147, CARPK, and ShanghaiTech crowd counting datasets demonstrate state-of-the-art accuracy and generalizability of the proposed method. Code is available: https://github.com/songrise/CLIP-Count.

Ye Liu, Huifang Li, Chao Hu et al.

The task of instance segmentation in remote sensing images, aiming at performing per-pixel labeling of objects at instance level, is of great importance for various civil applications. Despite previous successes, most existing instance segmentation methods designed for natural images encounter sharp performance degradations when they are directly applied to top-view remote sensing images. Through careful analysis, we observe that the challenges mainly come from the lack of discriminative object features due to severe scale variations, low contrasts, and clustered distributions. In order to address these problems, a novel context aggregation network (CATNet) is proposed to improve the feature extraction process. The proposed model exploits three lightweight plug-and-play modules, namely dense feature pyramid network (DenseFPN), spatial context pyramid (SCP), and hierarchical region of interest extractor (HRoIE), to aggregate global visual context at feature, spatial, and instance domains, respectively. DenseFPN is a multi-scale feature propagation module that establishes more flexible information flows by adopting inter-level residual connections, cross-level dense connections, and feature re-weighting strategy. Leveraging the attention mechanism, SCP further augments the features by aggregating global spatial context into local regions. For each instance, HRoIE adaptively generates RoI features for different downstream tasks. Extensive evaluations of the proposed scheme on iSAID, DIOR, NWPU VHR-10, and HRSID datasets demonstrate that the proposed approach outperforms state-of-the-arts under similar computational costs. Source code and pre-trained models are available at https://github.com/yeliudev/CATNet.

Rui Zhu, Bingchen Zhao, Jingen Liu et al.

Contrastive learning, which aims at minimizing the distance between positive pairs while maximizing that of negative ones, has been widely and successfully applied in unsupervised feature learning, where the design of positive and negative (pos/neg) pairs is one of its keys. In this paper, we attempt to devise a feature-level data manipulation, differing from data augmentation, to enhance the generic contrastive self-supervised learning. To this end, we first design a visualization scheme for pos/neg score (Pos/neg score indicates cosine similarity of pos/neg pair.) distribution, which enables us to analyze, interpret and understand the learning process. To our knowledge, this is the first attempt of its kind. More importantly, leveraging this tool, we gain some significant observations, which inspire our novel Feature Transformation proposals including the extrapolation of positives. This operation creates harder positives to boost the learning because hard positives enable the model to be more view-invariant. Besides, we propose the interpolation among negatives, which provides diversified negatives and makes the model more discriminative. It is the first attempt to deal with both challenges simultaneously. Experiment results show that our proposed Feature Transformation can improve at least 6.0% accuracy on ImageNet-100 over MoCo baseline, and about 2.0% accuracy on ImageNet-1K over the MoCoV2 baseline. Transferring to the downstream tasks successfully demonstrate our model is less task-bias. Visualization tools and codes https://github.com/DTennant/CL-Visualizing-Feature-Transformation .

Jiaxuan Li, Qing Xu, Xiangjian He et al.

Masked Autoencoders (MAE) achieve self-supervised learning of image representations by randomly removing a portion of visual tokens and reconstructing the original image as a pretext task, thereby significantly enhancing pretraining efficiency and yielding excellent adaptability across downstream tasks. However, MAE and other MAE-style paradigms that adopt random masking generally require more pre-training epochs to maintain adaptability. Meanwhile, ViT in MAE suffers from inefficient parameter use due to fixed spatial resolution across layers. To overcome these limitations, we propose the Complementary Masked Autoencoders (CoMA), which employ a complementary masking strategy to ensure uniform sampling across all pixels, thereby improving effective learning of all features and enhancing the model's adaptability. Furthermore, we introduce DyViT, a hierarchical vision transformer that employs a Dynamic Multi-Window Self-Attention (DM-MSA), significantly reducing the parameters and FLOPs while improving fine-grained feature learning. Pre-trained on ImageNet-1K with CoMA, DyViT matches the downstream performance of MAE using only 12% of the pre-training epochs, demonstrating more effective learning. It also attains a 10% reduction in pre-training time per epoch, further underscoring its superior pre-training efficiency.

Rui Zhu, Yingwei Pan, Yehao Li et al.

Diffusion Transformer (DiT) has emerged as the new trend of generative diffusion models on image generation. In view of extremely slow convergence in typical DiT, recent breakthroughs have been driven by mask strategy that significantly improves the training efficiency of DiT with additional intra-image contextual learning. Despite this progress, mask strategy still suffers from two inherent limitations: (a) training-inference discrepancy and (b) fuzzy relations between mask reconstruction & generative diffusion process, resulting in sub-optimal training of DiT. In this work, we address these limitations by novelly unleashing the self-supervised discrimination knowledge to boost DiT training. Technically, we frame our DiT in a teacher-student manner. The teacher-student discriminative pairs are built on the diffusion noises along the same Probability Flow Ordinary Differential Equation (PF-ODE). Instead of applying mask reconstruction loss over both DiT encoder and decoder, we decouple DiT encoder and decoder to separately tackle discriminative and generative objectives. In particular, by encoding discriminative pairs with student and teacher DiT encoders, a new discriminative loss is designed to encourage the inter-image alignment in the self-supervised embedding space. After that, student samples are fed into student DiT decoder to perform the typical generative diffusion task. Extensive experiments are conducted on ImageNet dataset, and our method achieves a competitive balance between training cost and generative capacity.

Binyuan Huang, Yuqing Wen, Yucheng Zhao et al.

Autonomous driving progress relies on large-scale annotated datasets. In this work, we explore the potential of generative models to produce vast quantities of freely-labeled data for autonomous driving applications and present SubjectDrive, the first model proven to scale generative data production in a way that could continuously improve autonomous driving applications. We investigate the impact of scaling up the quantity of generative data on the performance of downstream perception models and find that enhancing data diversity plays a crucial role in effectively scaling generative data production. Therefore, we have developed a novel model equipped with a subject control mechanism, which allows the generative model to leverage diverse external data sources for producing varied and useful data. Extensive evaluations confirm SubjectDrive's efficacy in generating scalable autonomous driving training data, marking a significant step toward revolutionizing data production methods in this field.

Zuyao Chen, Jinlin Wu, Zhen Lei et al.

Training Scene Graph Generation (SGG) models with natural language captions has become increasingly popular due to the abundant, cost-effective, and open-world generalization supervision signals that natural language offers. However, such unstructured caption data and its processing pose significant challenges in learning accurate and comprehensive scene graphs. The challenges can be summarized as three aspects: 1) traditional scene graph parsers based on linguistic representation often fail to extract meaningful relationship triplets from caption data. 2) grounding unlocalized objects of parsed triplets will meet ambiguity issues in visual-language alignment. 3) caption data typically are sparse and exhibit bias to partial observations of image content. Aiming to address these problems, we propose a divide-and-conquer strategy with a novel framework named \textit{GPT4SGG}, to obtain more accurate and comprehensive scene graph signals. This framework decomposes a complex scene into a bunch of simple regions, resulting in a set of region-specific narratives. With these region-specific narratives (partial observations) and a holistic narrative (global observation) for an image, a large language model (LLM) performs the relationship reasoning to synthesize an accurate and comprehensive scene graph. Experimental results demonstrate \textit{GPT4SGG} significantly improves the performance of SGG models trained on image-caption data, in which the ambiguity issue and long-tail bias have been well-handled with more accurate and comprehensive scene graphs.

Zuyao Chen, Jinlin Wu, Zhen Lei et al.

While text-to-image generation has been extensively studied, generating images from scene graphs remains relatively underexplored, primarily due to challenges in accurately modeling spatial relationships and object interactions. To fill this gap, we introduce Scene-Bench, a comprehensive benchmark designed to evaluate and enhance the factual consistency in generating natural scenes. Scene-Bench comprises MegaSG, a large-scale dataset of one million images annotated with scene graphs, facilitating the training and fair comparison of models across diverse and complex scenes. Additionally, we propose SGScore, a novel evaluation metric that leverages chain-of-thought reasoning capabilities of multimodal large language models (LLMs) to assess both object presence and relationship accuracy, offering a more effective measure of factual consistency than traditional metrics like FID and CLIPScore. Building upon this evaluation framework, we develop a scene graph feedback pipeline that iteratively refines generated images by identifying and correcting discrepancies between the scene graph and the image. Extensive experiments demonstrate that Scene-Bench provides a more comprehensive and effective evaluation framework compared to existing benchmarks, particularly for complex scene generation. Furthermore, our feedback strategy significantly enhances the factual consistency of image generation models, advancing the field of controllable image generation.

Tianyu Zhang, Dong Liu, Chang Wen Chen

Ultra-low bitrate image compression (below 0.05 bits per pixel) is increasingly critical for bandwidth-constrained and computation-limited encoding scenarios such as edge devices. Existing frameworks typically rely on large pretrained encoders (e.g., VAEs or tokenizer-based models) and perform transform coding within their generative latent space. While these approaches achieve impressive perceptual fidelity, their reliance on heavy encoder networks makes them unsuitable for deployment on weak sender devices. In this work, we explore the feasibility of applying shallow encoders for ultra-low bitrate compression and propose a novel Asymmetric Extreme Image Compression (AEIC) framework that pursues simultaneously encoding simplicity and decoding quality. Specifically, AEIC employs moderate or even shallow encoder networks, while leveraging an one-step diffusion decoder to maintain high-fidelity and high-realism reconstructions under extreme bitrates. To further enhance the efficiency of shallow encoders, we design a dual-side feature distillation scheme that transfers knowledge from AEIC with moderate encoders to its shallow encoder variants. Experiments demonstrate that AEIC not only outperforms existing methods on rate-distortion-perception performance at ultra-low bitrates, but also delivers exceptional encoding efficiency for 35.8 FPS on 1080P input images, while maintaining competitive decoding speed compared to existing methods.

Yuang Chen, Fengqian Guo, Chang Wu et al.

In the Industrial Internet of Things (IIoT), the frequent transmission of large amounts of data over wireless networks should meet the stringent timeliness requirements. Particularly, the freshness of packet status updates has a significant impact on the system performance. In this paper, we propose an age-of-information (AoI)-aware multi-base station (BS) real-time monitoring framework to support extensive IIoT deployments. To meet the freshness requirements of IIoT, we formulate a joint task offloading and resource allocation optimization problem with the goal of minimizing long-term average AoI. Tackling the core challenges of combinatorial explosion in multi-BS decision spaces and the stochastic dynamics of IIoT systems is crucial, as these factors render traditional optimization methods intractable. Firstly, an innovative branching-based Dueling Double Deep Q-Network (Branching-D3QN) algorithm is proposed to effectively implement task offloading, which optimizes the convergence performance by reducing the action space complexity from exponential to linear levels. Then, an efficient optimization solution to resource allocation is proposed by proving the semi-definite property of the Hessian matrix of bandwidth and computation resources. Finally, we propose an iterative optimization algorithm for efficient joint task offloading and resource allocation to achieve optimal average AoI performance. Extensive simulations demonstrate that our proposed Branching-D3QN algorithm outperforms both state-of-the-art DRL methods and classical heuristics, achieving up to a 75% enhanced convergence speed and at least a 22% reduction in the long-term average AoI.

Ye Liu, Zongyang Ma, Junfu Pu et al.

Recent advances in Large Multi-modal Models (LMMs) have demonstrated their remarkable success as general-purpose multi-modal assistants, with particular focuses on holistic image- and video-language understanding. Conversely, less attention has been given to scaling fine-grained pixel-level understanding capabilities, where the models are expected to realize pixel-level alignment between visual signals and language semantics. Some previous studies have applied LMMs to related tasks such as region-level captioning and referring expression segmentation. However, these models are limited to performing either referring or segmentation tasks independently and fail to integrate these fine-grained perception capabilities into visual reasoning. To bridge this gap, we propose UniPixel, a large multi-modal model capable of flexibly comprehending visual prompt inputs and generating mask-grounded responses. Our model distinguishes itself by seamlessly integrating pixel-level perception with general visual understanding capabilities. Specifically, UniPixel processes visual prompts and generates relevant masks on demand, and performs subsequent reasoning conditioning on these intermediate pointers during inference, thereby enabling fine-grained pixel-level reasoning. The effectiveness of our approach has been verified on 10 benchmarks across a diverse set of tasks, including pixel-level referring/segmentation and object-centric understanding in images/videos. A novel PixelQA task that jointly requires referring, segmentation, and question answering is also designed to verify the flexibility of our method.

Nannan Wu, Zengqiang Yan, Nong Sang et al.

Training a model that effectively handles both common and rare data-i.e., achieving performance fairness-is crucial in federated learning (FL). While existing fair FL methods have shown effectiveness, they remain vulnerable to mislabeled data. Ensuring robustness in fair FL is therefore essential. However, fairness and robustness inherently compete, which causes robust strategies to hinder fairness. In this paper, we attribute this competition to the homogeneity in loss patterns exhibited by rare and mislabeled data clients, preventing existing loss-based fair and robust FL methods from effectively distinguishing and handling these two distinct client types. To address this, we propose performance-capacity analysis, which jointly considers model performance on each client and its capacity to handle the dataset, measured by loss and a newly introduced feature dispersion score. This allows mislabeled clients to be identified by their significantly deviated performance relative to capacity while preserving rare data clients. Building on this, we introduce FedPCA, an FL method that robustly achieves fairness. FedPCA first identifies mislabeled clients via a Gaussian Mixture Model on loss-dispersion pairs, then applies fairness and robustness strategies in global aggregation and local training by adjusting client weights and selectively using reliable data. Extensive experiments on three datasets demonstrate FedPCA's effectiveness in tackling this complex challenge. Code will be publicly available upon acceptance.

Xiaotong Yu, Ruihan Xie, Zhihe Zhao et al.

While we enjoy the richness and informativeness of multimodal data, it also introduces interference and redundancy of information. To achieve optimal domain interpretation with limited resources, we propose CSDNet, a lightweight \textbf{C}ross \textbf{S}hallow and \textbf{D}eep Perception \textbf{Net}work designed to integrate two modalities with less coherence, thereby discarding redundant information or even modality. We implement our CSDNet for Salient Object Detection (SOD) task in robotic perception. The proposed method capitalises on spatial information prescreening and implicit coherence navigation across shallow and deep layers of the depth-thermal (D-T) modality, prioritising integration over fusion to maximise the scene interpretation. To further refine the descriptive capabilities of the encoder for the less-known D-T modalities, we also propose SAMAEP to guide an effective feature mapping to the generalised feature space. Our approach is tested on the VDT-2048 dataset, leveraging the D-T modality outperforms those of SOTA methods using RGB-T or RGB-D modalities for the first time, achieves comparable performance with the RGB-D-T triple-modality benchmark method with 5.97 times faster at runtime and demanding 0.0036 times fewer FLOPs. Demonstrates the proposed CSDNet effectively integrates the information from the D-T modality. The code will be released upon acceptance.

Jiaming Shen, Kun Hu, Wei Bao et al.

Hand-drawn 2D animation workflow is typically initiated with the creation of sketch keyframes. Subsequent manual inbetweens are crafted for smoothness, which is a labor-intensive process and the prospect of automatic animation sketch interpolation has become highly appealing. Yet, common frame interpolation methods are generally hindered by two key issues: 1) limited texture and colour details in sketches, and 2) exaggerated alterations between two sketch keyframes. To overcome these issues, we propose a novel deep learning method - Sketch-Aware Interpolation Network (SAIN). This approach incorporates multi-level guidance that formulates region-level correspondence, stroke-level correspondence and pixel-level dynamics. A multi-stream U-Transformer is then devised to characterize sketch inbetweening patterns using these multi-level guides through the integration of self / cross-attention mechanisms. Additionally, to facilitate future research on animation sketch inbetweening, we constructed a large-scale dataset - STD-12K, comprising 30 sketch animation series in diverse artistic styles. Comprehensive experiments on this dataset convincingly show that our proposed SAIN surpasses the state-of-the-art interpolation methods.

Bruce X. B. Yu, Jianlong Chang, Haixin Wang et al.

Fine-tuning visual models has been widely shown promising performance on many downstream visual tasks. With the surprising development of pre-trained visual foundation models, visual tuning jumped out of the standard modus operandi that fine-tunes the whole pre-trained model or just the fully connected layer. Instead, recent advances can achieve superior performance than full-tuning the whole pre-trained parameters by updating far fewer parameters, enabling edge devices and downstream applications to reuse the increasingly large foundation models deployed on the cloud. With the aim of helping researchers get the full picture and future directions of visual tuning, this survey characterizes a large and thoughtful selection of recent works, providing a systematic and comprehensive overview of existing work and models. Specifically, it provides a detailed background of visual tuning and categorizes recent visual tuning techniques into five groups: prompt tuning, adapter tuning, parameter tuning, and remapping tuning. Meanwhile, it offers some exciting research directions for prospective pre-training and various interactions in visual tuning.

Tao Wei, Angelica I Aviles-Rivero, Shuo Wang et al.

The task of classifying mammograms is very challenging because the lesion is usually small in the high resolution image. The current state-of-the-art approaches for medical image classification rely on using the de-facto method for ConvNets - fine-tuning. However, there are fundamental differences between natural images and medical images, which based on existing evidence from the literature, limits the overall performance gain when designed with algorithmic approaches. In this paper, we propose to go beyond fine-tuning by introducing a novel framework called MorphHR, in which we highlight a new transfer learning scheme. The idea behind the proposed framework is to integrate function-preserving transformations, for any continuous non-linear activation neurons, to internally regularise the network for improving mammograms classification. The proposed solution offers two major advantages over the existing techniques. Firstly and unlike fine-tuning, the proposed approach allows for modifying not only the last few layers but also several of the first ones on a deep ConvNet. By doing this, we can design the network front to be suitable for learning domain specific features. Secondly, the proposed scheme is scalable to hardware. Therefore, one can fit high resolution images on standard GPU memory. We show that by using high resolution images, one prevents losing relevant information. We demonstrate, through numerical and visual experiments, that the proposed approach yields to a significant improvement in the classification performance over state-of-the-art techniques, and is indeed on a par with radiology experts. Moreover and for generalisation purposes, we show the effectiveness of the proposed learning scheme on another large dataset, the ChestX-ray14, surpassing current state-of-the-art techniques.

Lifang Wu, Zhou Yang, Qi Wang et al.

Many semantic events in team sport activities e.g. basketball often involve both group activities and the outcome (score or not). Motion patterns can be an effective means to identify different activities. Global and local motions have their respective emphasis on different activities, which are difficult to capture from the optical flow due to the mixture of global and local motions. Hence it calls for a more effective way to separate the global and local motions. When it comes to the specific case for basketball game analysis, the successful score for each round can be reliably detected by the appearance variation around the basket. Based on the observations, we propose a scheme to fuse global and local motion patterns (MPs) and key visual information (KVI) for semantic event recognition in basketball videos. Firstly, an algorithm is proposed to estimate the global motions from the mixed motions based on the intrinsic property of camera adjustments. And the local motions could be obtained from the mixed and global motions. Secondly, a two-stream 3D CNN framework is utilized for group activity recognition over the separated global and local motion patterns. Thirdly, the basket is detected and its appearance features are extracted through a CNN structure. The features are utilized to predict the success or failure. Finally, the group activity recognition and success/failure prediction results are integrated using the kronecker product for event recognition. Experiments on NCAA dataset demonstrate that the proposed method obtains state-of-the-art performance.

Yu Liu, Li Deng, Jianshu Chen et al.

Many important classification problems, such as object classification, speech recognition, and machine translation, have been tackled by the supervised learning paradigm in the past, where training corpora of parallel input-output pairs are required with high cost. To remove the need for the parallel training corpora has practical significance for real-world applications, and it is one of the main goals of unsupervised learning. Recently, encouraging progress in unsupervised learning for solving such classification problems has been made and the nature of the challenges has been clarified. In this article, we review this progress and disseminate a class of promising new methods to facilitate understanding the methods for machine learning researchers. In particular, we emphasize the key information that enables the success of unsupervised learning - the sequential statistics as the distributional prior in the labels. Exploitation of such sequential statistics makes it possible to estimate parameters of classifiers without the need of paired input-output data. In this paper, we first introduce the concept of Caesar Cipher and its decryption, which motivated the construction of the novel loss function for unsupervised learning we use throughout the paper. Then we use a simple but representative binary classification task as an example to derive and describe the unsupervised learning algorithm in a step-by-step, easy-to-understand fashion. We include two cases, one with Bigram language model as the sequential statistics for use in unsupervised parameter estimation, and another with a simpler Unigram language model. For both cases, detailed derivation steps for the learning algorithm are included. Further, a summary table compares computational steps of the two cases in executing the unsupervised learning algorithm for learning binary classifiers.

Guo-Jun Qi, Liheng Zhang, Chang Wen Chen et al.

The learning of Transformation-Equivariant Representations (TERs), which is introduced by Hinton et al. \cite{hinton2011transforming}, has been considered as a principle to reveal visual structures under various transformations. It contains the celebrated Convolutional Neural Networks (CNNs) as a special case that only equivary to the translations. In contrast, we seek to train TERs for a generic class of transformations and train them in an {\em unsupervised} fashion. To this end, we present a novel principled method by Autoencoding Variational Transformations (AVT), compared with the conventional approach to autoencoding data. Formally, given transformed images, the AVT seeks to train the networks by maximizing the mutual information between the transformations and representations. This ensures the resultant TERs of individual images contain the {\em intrinsic} information about their visual structures that would equivary {\em extricably} under various transformations in a generalized {\em nonlinear} case. Technically, we show that the resultant optimization problem can be efficiently solved by maximizing a variational lower-bound of the mutual information. This variational approach introduces a transformation decoder to approximate the intractable posterior of transformations, resulting in an autoencoding architecture with a pair of the representation encoder and the transformation decoder. Experiments demonstrate the proposed AVT model sets a new record for the performances on unsupervised tasks, greatly closing the performance gap to the supervised models.

Lifang Wu, Zhou Yang, Jiaoyu He et al.

In multi-person videos, especially team sport videos, a semantic event is usually represented as a confrontation between two teams of players, which can be represented as collective motion. In broadcast basketball videos, specific camera motions are used to present specific events. Therefore, a semantic event in broadcast basketball videos is closely related to both the global motion (camera motion) and the collective motion. A semantic event in basketball videos can be generally divided into three stages: pre-event, event occurrence (event-occ), and post-event. In this paper, we propose an ontology-based global and collective motion pattern (On_GCMP) algorithm for basketball event classification. First, a two-stage GCMP based event classification scheme is proposed. The GCMP is extracted using optical flow. The two-stage scheme progressively combines a five-class event classification algorithm on event-occs and a two-class event classification algorithm on pre-events. Both algorithms utilize sequential convolutional neural networks (CNNs) and long short-term memory (LSTM) networks to extract the spatial and temporal features of GCMP for event classification. Second, we utilize post-event segments to predict success/failure using deep features of images in the video frames (RGB_DF_VF) based algorithms. Finally the event classification results and success/failure classification results are integrated to obtain the final results. To evaluate the proposed scheme, we collected a new dataset called NCAA+, which is automatically obtained from the NCAA dataset by extending the fixed length of video clips forward and backward of the corresponding semantic events. The experimental results demonstrate that the proposed scheme achieves the mean average precision of 58.10% on NCAA+. It is higher by 6.50% than state-of-the-art on NCAA.

Xiaoda Jiang, Hancheng Lu, Chang Wen Chen et al.

Non-orthogonal multiple access (NOMA) has shown potential for scalable multicast of video data. However, one key drawback for NOMA-based video multicast is the limited number of layers allowed by the embedded successive interference cancellation algorithm, failing to meet satisfaction of heterogeneous receivers. We propose a novel receiver-driven superposed video multicast (Supcast) scheme by integrating Softcast, an analog-like transmission scheme, into the NOMA-based system to achieve high bandwidth efficiency as well as gradual decoding quality proportional to channel conditions at receivers. Although Softcast allows gradual performance by directly transmitting power-scaled transformation coefficients of frames, it suffers performance degradation due to discarding coefficients under insufficient bandwidth and its power allocation strategy cannot be directly applied in NOMA due to interference. In Supcast, coefficients are grouped into chunks, which are basic units for power allocation and superposition scheduling. By bisecting chunks into base-layer chunks and enhanced-layer chunks, the joint power allocation and chunk scheduling is formulated as a distortion minimization problem. A two-stage power allocation strategy and a near-optimal low-complexity algorithm for chunk scheduling based on the matching theory are proposed. Simulation results have shown the advantage of Supcast against Softcast as well as the reference scheme in NOMA under various practical scenarios.

Shuang Ma, Jianlong Fu, Chang Wen Chen et al.

Unsupervised image translation, which aims in translating two independent sets of images, is challenging in discovering the correct correspondences without paired data. Existing works build upon Generative Adversarial Network (GAN) such that the distribution of the translated images are indistinguishable from the distribution of the target set. However, such set-level constraints cannot learn the instance-level correspondences (e.g. aligned semantic parts in object configuration task). This limitation often results in false positives (e.g. geometric or semantic artifacts), and further leads to mode collapse problem. To address the above issues, we propose a novel framework for instance-level image translation by Deep Attention GAN (DA-GAN). Such a design enables DA-GAN to decompose the task of translating samples from two sets into translating instances in a highly-structured latent space. Specifically, we jointly learn a deep attention encoder, and the instancelevel correspondences could be consequently discovered through attending on the learned instance pairs. Therefore, the constraints could be exploited on both set-level and instance-level. Comparisons against several state-ofthe- arts demonstrate the superiority of our approach, and the broad application capability, e.g, pose morphing, data augmentation, etc., pushes the margin of domain translation problem.

Jing Zhang, Yang Cao, Yang Wang et al.

Modeling statistical regularity plays an essential role in ill-posed image processing problems. Recently, deep learning based methods have been presented to implicitly learn statistical representation of pixel distributions in natural images and leverage it as a constraint to facilitate subsequent tasks, such as color constancy and image dehazing. However, the existing CNN architecture is prone to variability and diversity of pixel intensity within and between local regions, which may result in inaccurate statistical representation. To address this problem, this paper presents a novel fully point-wise CNN architecture for modeling statistical regularities in natural images. Specifically, we propose to randomly shuffle the pixels in the origin images and leverage the shuffled image as input to make CNN more concerned with the statistical properties. Moreover, since the pixels in the shuffled image are independent identically distributed, we can replace all the large convolution kernels in CNN with point-wise ($1*1$) convolution kernels while maintaining the representation ability. Experimental results on two applications: color constancy and image dehazing, demonstrate the superiority of our proposed network over the existing architectures, i.e., using 1/10$\sim$1/100 network parameters and computational cost while achieving comparable performance.

Xiaoda Jiang, Hancheng Lu, Chang Wen Chen

Recently, non-orthogonal multiple access (NOMA) has been proposed to achieve higher spectral efficiency over conventional orthogonal multiple access. Although it has the potential to meet increasing demands of video services, it is still challenging to provide high performance video streaming. In this research, we investigate, for the first time, a multi-user NOMA system design for video transmission. Various NOMA systems have been proposed for data transmission in terms of throughput or reliability. However, the perceived quality, or the quality-of-experience of users, is more critical for video transmission. Based on this observation, we design a quality-driven scalable video transmission framework with cross-layer support for multi-user NOMA. To enable low complexity multi-user NOMA operations, a novel user grouping strategy is proposed. The key features in the proposed framework include the integration of the quality model for encoded video with the physical layer model for NOMA transmission, and the formulation of multi-user NOMA-based video transmission as a quality-driven power allocation problem. As the problem is non-concave, a global optimal algorithm based on the hidden monotonic property and a suboptimal algorithm with polynomial time complexity are developed. Simulation results show that the proposed multi-user NOMA system outperforms existing schemes in various video delivery scenarios.

Shuang Ma, Jing Liu, Chang Wen Chen

Deep convolutional neural networks (CNN) have recently been shown to generate promising results for aesthetics assessment. However, the performance of these deep CNN methods is often compromised by the constraint that the neural network only takes the fixed-size input. To accommodate this requirement, input images need to be transformed via cropping, warping, or padding, which often alter image composition, reduce image resolution, or cause image distortion. Thus the aesthetics of the original images is impaired because of potential loss of fine grained details and holistic image layout. However, such fine grained details and holistic image layout is critical for evaluating an image's aesthetics. In this paper, we present an Adaptive Layout-Aware Multi-Patch Convolutional Neural Network (A-Lamp CNN) architecture for photo aesthetic assessment. This novel scheme is able to accept arbitrary sized images, and learn from both fined grained details and holistic image layout simultaneously. To enable training on these hybrid inputs, we extend the method by developing a dedicated double-subnet neural network structure, i.e. a Multi-Patch subnet and a Layout-Aware subnet. We further construct an aggregation layer to effectively combine the hybrid features from these two subnets. Extensive experiments on the large-scale aesthetics assessment benchmark (AVA) demonstrate significant performance improvement over the state-of-the-art in photo aesthetic assessment.

Tao Wei, Changhu Wang, Chang Wen Chen

In this work we study the problem of network morphism, an effective learning scheme to morph a well-trained neural network to a new one with the network function completely preserved. Different from existing work where basic morphing types on the layer level were addressed, we target at the central problem of network morphism at a higher level, i.e., how a convolutional layer can be morphed into an arbitrary module of a neural network. To simplify the representation of a network, we abstract a module as a graph with blobs as vertices and convolutional layers as edges, based on which the morphing process is able to be formulated as a graph transformation problem. Two atomic morphing operations are introduced to compose the graphs, based on which modules are classified into two families, i.e., simple morphable modules and complex modules. We present practical morphing solutions for both of these two families, and prove that any reasonable module can be morphed from a single convolutional layer. Extensive experiments have been conducted based on the state-of-the-art ResNet on benchmark datasets, and the effectiveness of the proposed solution has been verified.

Lei Yu, Houqiang Li, Chang Wen Chen

In literature, different common informations were defined by Gács and Körner, by Wyner, and by Kumar, Li, and Gamal, respectively. In this paper, we define two generalized versions of common informations, named approximate and exact information-correlation functions, by exploiting the conditional maximal correlation as a commonness or privacy measure. These two generalized common informations encompass the notions of Gács-Körner's, Wyner's, and Kumar-Li-Gamal's common informations as special cases. Furthermore, to give operational characterizations of these two generalized common informations, we also study the problems of private sources synthesis and common information extraction, and show that the information-correlation functions are equal to the minimum rates of commonness needed to ensure that some conditional maximal correlation constraints are satisfied for the centralized setting versions of these problems. As a byproduct, the conditional maximal correlation has been studied as well.