Chunjie Zhang

Man Liu, Feng Li, Chunjie Zhang et al.

Generalized Zero-Shot Learning (GZSL) identifies unseen categories by knowledge transferred from the seen domain, relying on the intrinsic interactions between visual and semantic information. Prior works mainly localize regions corresponding to the sharing attributes. When various visual appearances correspond to the same attribute, the sharing attributes inevitably introduce semantic ambiguity, hampering the exploration of accurate semantic-visual interactions. In this paper, we deploy the dual semantic-visual transformer module (DSVTM) to progressively model the correspondences between attribute prototypes and visual features, constituting a progressive semantic-visual mutual adaption (PSVMA) network for semantic disambiguation and knowledge transferability improvement. Specifically, DSVTM devises an instance-motivated semantic encoder that learns instance-centric prototypes to adapt to different images, enabling the recast of the unmatched semantic-visual pair into the matched one. Then, a semantic-motivated instance decoder strengthens accurate cross-domain interactions between the matched pair for semantic-related instance adaption, encouraging the generation of unambiguous visual representations. Moreover, to mitigate the bias towards seen classes in GZSL, a debiasing loss is proposed to pursue response consistency between seen and unseen predictions. The PSVMA consistently yields superior performances against other state-of-the-art methods. Code will be available at: https://github.com/ManLiuCoder/PSVMA.

Hongguang Zhu, Yunchao Wei, Xiaodan Liang et al.

Vision-Language Pretraining (VLP) has shown impressive results on diverse downstream tasks by offline training on large-scale datasets. Regarding the growing nature of real-world data, such an offline training paradigm on ever-expanding data is unsustainable, because models lack the continual learning ability to accumulate knowledge constantly. However, most continual learning studies are limited to uni-modal classification and existing multi-modal datasets cannot simulate continual non-stationary data stream scenarios. To support the study of Vision-Language Continual Pretraining (VLCP), we first contribute a comprehensive and unified benchmark dataset P9D which contains over one million product image-text pairs from 9 industries. The data from each industry as an independent task supports continual learning and conforms to the real-world long-tail nature to simulate pretraining on web data. We comprehensively study the characteristics and challenges of VLCP, and propose a new algorithm: Compatible momentum contrast with Topology Preservation, dubbed CTP. The compatible momentum model absorbs the knowledge of the current and previous-task models to flexibly update the modal feature. Moreover, Topology Preservation transfers the knowledge of embedding across tasks while preserving the flexibility of feature adjustment. The experimental results demonstrate our method not only achieves superior performance compared with other baselines but also does not bring an expensive training burden. Dataset and codes are available at https://github.com/KevinLight831/CTP.

Anqi Li, Feng Li, Jiaxin Han et al.

Learned image compression (LIC) methods have experienced significant progress during recent years. However, these methods are primarily dedicated to optimizing the rate-distortion (R-D) performance at medium and high bitrates (> 0.1 bits per pixel (bpp)), while research on extremely low bitrates is limited. Besides, existing methods fail to explicitly explore the image structure and texture components crucial for image compression, treating them equally alongside uninformative components in networks. This can cause severe perceptual quality degradation, especially under low-bitrate scenarios. In this work, inspired by the success of pre-trained masked autoencoders (MAE) in many downstream tasks, we propose to rethink its mask sampling strategy from structure and texture perspectives for high redundancy reduction and discriminative feature representation, further unleashing the potential of LIC methods. Therefore, we present a dual-adaptive masking approach (DA-Mask) that samples visible patches based on the structure and texture distributions of original images. We combine DA-Mask and pre-trained MAE in masked image modeling (MIM) as an initial compressor that abstracts informative semantic context and texture representations. Such a pipeline can well cooperate with LIC networks to achieve further secondary compression while preserving promising reconstruction quality. Consequently, we propose a simple yet effective masked compression model (MCM), the first framework that unifies MIM and LIC end-to-end for extremely low-bitrate image compression. Extensive experiments have demonstrated that our approach outperforms recent state-of-the-art methods in R-D performance, visual quality, and downstream applications, at very low bitrates. Our code is available at https://github.com/lianqi1008/MCM.git.

Panyu Zhai, Yanwu Yang, Chunjie Zhang

As a prevalent problem in online advertising, CTR prediction has attracted plentiful attention from both academia and industry. Recent studies have been reported to establish CTR prediction models in the graph neural networks (GNNs) framework. However, most of GNNs-based models handle feature interactions in a complete graph, while ignoring causal relationships among features, which results in a huge drop in the performance on out-of-distribution data. This paper is dedicated to developing a causality-based CTR prediction model in the GNNs framework (Causal-GNN) integrating representations of feature graph, user graph and ad graph in the context of online advertising. In our model, a structured representation learning method (GraphFwFM) is designed to capture high-order representations on feature graph based on causal discovery among field features in gated graph neural networks (GGNNs), and GraphSAGE is employed to obtain graph representations of users and ads. Experiments conducted on three public datasets demonstrate the superiority of Causal-GNN in AUC and Logloss and the effectiveness of GraphFwFM in capturing high-order representations on causal feature graph.

Peixuan Zhang, Chang Zhou, Ziyuan Zhang et al.

The surging demand for adapting long-form cinematic content into short videos has motivated the need for versatile automatic video compilation systems. However, existing compilation methods are limited to predefined tasks, and the community lacks a comprehensive benchmark to evaluate the cinematic compilation. To address this, we introduce CineBench, the first benchmark for instruction-driven cinematic video compilation, featuring diverse user instructions and high-quality ground-truth compilations annotated by professional editors. To overcome contextual collapse and temporal fragmentation, we present CineAgents, a multi-agent system that reformulates cinematic video compilation into ``design-and-compose'' paradigm. CineAgents performs script reverse-engineering to construct a hierarchical narrative memory to provide multi-level context and employs an iterative narrative planning process that refines a creative blueprint into a final compiled script. Extensive experiments demonstrate that CineAgents significantly outperforms existing methods, generating compilations with superior narrative coherence and logical coherence.

Qiuhai Yan, Aiwen Jiang, Kang Chen et al.

Image restoration has always been a cutting-edge topic in the academic and industrial fields of computer vision. Since degradation signals are often random and diverse, "all-in-one" models that can do blind image restoration have been concerned in recent years. Early works require training specialized headers and tails to handle each degradation of concern, which are manually cumbersome. Recent works focus on learning visual prompts from data distribution to identify degradation type. However, the prompts employed in most of models are non-text, lacking sufficient emphasis on the importance of human-in-the-loop. In this paper, an effective textual prompt guided image restoration model has been proposed. In this model, task-specific BERT is fine-tuned to accurately understand user's instructions and generating textual prompt guidance. Depth-wise multi-head transposed attentions and gated convolution modules are designed to bridge the gap between textual prompts and visual features. The proposed model has innovatively introduced semantic prompts into low-level visual domain. It highlights the potential to provide a natural, precise, and controllable way to perform image restoration tasks. Extensive experiments have been done on public denoising, dehazing and deraining datasets. The experiment results demonstrate that, compared with popular state-of-the-art methods, the proposed model can obtain much more superior performance, achieving accurate recognition and removal of degradation without increasing model's complexity. Related source codes and data will be publicly available on github site https://github.com/MoTong-AI-studio/TextPromptIR.

Henghui Du, Guangyao Li, Chang Zhou et al.

In recent years, numerous tasks have been proposed to encourage model to develop specified capability in understanding audio-visual scene, primarily categorized into temporal localization, spatial localization, spatio-temporal reasoning, and pixel-level understanding. Instead, human possesses a unified understanding ability for diversified tasks. Therefore, designing an audio-visual model with general capability to unify these tasks is of great value. However, simply joint training for all tasks can lead to interference due to the heterogeneity of audiovisual data and complex relationship among tasks. We argue that this problem can be solved through explicit cooperation among tasks. To achieve this goal, we propose a unified learning method which achieves explicit inter-task cooperation from both the perspectives of data and model thoroughly. Specifically, considering the labels of existing datasets are simple words, we carefully refine these datasets and construct an Audio-Visual Unified Instruction-tuning dataset with Explicit reasoning process (AV-UIE), which clarifies the cooperative relationship among tasks. Subsequently, to facilitate concrete cooperation in learning stage, an interaction-aware LoRA structure with multiple LoRA heads is designed to learn different aspects of audiovisual data interaction. By unifying the explicit cooperation across the data and model aspect, our method not only surpasses existing unified audio-visual model on multiple tasks, but also outperforms most specialized models for certain tasks. Furthermore, we also visualize the process of explicit cooperation and surprisingly find that each LoRA head has certain audio-visual understanding ability. Code and dataset: https://github.com/GeWu-Lab/Crab

Henghui Du, Chunjie Zhang, Xi Chen et al.

Long Video Question-Answering (LVQA) presents a significant challenge for Multi-modal Large Language Models (MLLMs) due to immense context and overloaded information, which could also lead to prohibitive memory consumption. While existing methods attempt to address these issues by reducing visual tokens or extending model's context length, they may miss useful information or take considerable computation. In fact, when answering given questions, only a small amount of crucial information is required. Therefore, we propose an efficient question-aware memory mechanism, enabling MLLMs to recurrently seek these critical clues. Our approach, named VideoDetective, simplifies this task by iteratively processing video sub-segments. For each sub-segment, a question-aware compression strategy is employed by introducing a few special memory tokens to achieve purposefully compression. This allows models to effectively seek critical clues while reducing visual tokens. Then, due to history context could have a significant impact, we recurrently aggregate and store these memory tokens to update history context, which would be reused for subsequent sub-segments. Furthermore, to more effectively measure model's long video understanding ability, we introduce GLVC (Grounding Long Video Clues), a long video question-answering dataset, which features grounding critical and concrete clues scattered throughout entire videos. Experimental results demonstrate our method enables MLLMs with limited context length of 32K to efficiently process 100K tokens (3600 frames, an hour-long video sampled at 1fps), requiring only 2 minutes and 37GB GPU memory usage. Evaluation results across multiple long video benchmarks illustrate our method can more effectively seek critical clues from massive information.

Man Liu, Huihui Bai, Feng Li et al.

Generalized zero-shot learning (GZSL) endeavors to identify the unseen categories using knowledge from the seen domain, necessitating the intrinsic interactions between the visual features and attribute semantic features. However, GZSL suffers from insufficient visual-semantic correspondences due to the attribute diversity and instance diversity. Attribute diversity refers to varying semantic granularity in attribute descriptions, ranging from low-level (specific, directly observable) to high-level (abstract, highly generic) characteristics. This diversity challenges the collection of adequate visual cues for attributes under a uni-granularity. Additionally, diverse visual instances corresponding to the same sharing attributes introduce semantic ambiguity, leading to vague visual patterns. To tackle these problems, we propose a multi-granularity progressive semantic-visual mutual adaption (PSVMA+) network, where sufficient visual elements across granularity levels can be gathered to remedy the granularity inconsistency. PSVMA+ explores semantic-visual interactions at different granularity levels, enabling awareness of multi-granularity in both visual and semantic elements. At each granularity level, the dual semantic-visual transformer module (DSVTM) recasts the sharing attributes into instance-centric attributes and aggregates the semantic-related visual regions, thereby learning unambiguous visual features to accommodate various instances. Given the diverse contributions of different granularities, PSVMA+ employs selective cross-granularity learning to leverage knowledge from reliable granularities and adaptively fuses multi-granularity features for comprehensive representations. Experimental results demonstrate that PSVMA+ consistently outperforms state-of-the-art methods.

Shibo Yao, Renshuai Tao, Xiaolong Zheng et al.

Recent deepfake detection studies often treat unseen sample detection as a ``zero-shot" task, training on images generated by known models but generalizing to unknown ones. A key real-world challenge arises when a model performs poorly on unknown samples, yet these samples remain available for analysis. This highlights that it should be approached as a ``few-shot" task, where effectively utilizing a small number of samples can lead to significant improvement. Unlike typical few-shot tasks focused on semantic understanding, deepfake detection prioritizes image realism, which closely mirrors real-world distributions. In this work, we propose the Few-shot Training-free Network (FTNet) for real-world few-shot deepfake detection. Simple yet effective, FTNet differs from traditional methods that rely on large-scale known data for training. Instead, FTNet uses only one fake samplefrom an evaluation set, mimicking the scenario where new samples emerge in the real world and can be gathered for use, without any training or parameter updates. During evaluation, each test sample is compared to the known fake and real samples, and it is classified based on the category of the nearest sample. We conduct a comprehensive analysis of AI-generated images from 29 different generative models and achieve a new SoTA performance, with an average improvement of 8.7\% compared to existing methods. This work introduces a fresh perspective on real-world deepfake detection: when the model struggles to generalize on a few-shot sample, leveraging the failed samples leads to better performance.

Zhiqiang Yang, Renshuai Tao, Xiaolong Zheng et al.

Existing deepfake detection methods heavily depend on labeled training data. However, as AI-generated content becomes increasingly realistic, even \textbf{human annotators struggle to distinguish} between deepfakes and authentic images. This makes the labeling process both time-consuming and less reliable. Specifically, there is a growing demand for approaches that can effectively utilize large-scale unlabeled data from online social networks. Unlike typical unsupervised learning tasks, where categories are distinct, AI-generated faces closely mimic real image distributions and share strong similarities, causing performance drop in conventional strategies. In this paper, we introduce the Dual-Path Guidance Network (DPGNet), to tackle two key challenges: (1) bridging the domain gap between faces from different generation models, and (2) utilizing unlabeled image samples. The method features two core modules: text-guided cross-domain alignment, which uses learnable prompts to unify visual and textual embeddings into a domain-invariant feature space, and curriculum-driven pseudo label generation, which dynamically exploit more informative unlabeled samples. To prevent catastrophic forgetting, we also facilitate bridging between domains via cross-domain knowledge distillation. Extensive experiments on \textbf{11 popular datasets}, show that DPGNet outperforms SoTA approaches by \textbf{6.3\%}, highlighting its effectiveness in leveraging unlabeled data to address the annotation challenges posed by the increasing realism of deepfakes.

Man Liu, Huihui Bai, Feng Li et al.

Zero-shot learning (ZSL) endeavors to transfer knowledge from seen categories to recognize unseen categories, which mostly relies on the semantic-visual interactions between image and attribute tokens. Recently, prompt learning has emerged in ZSL and demonstrated significant potential as it allows the zero-shot transfer of diverse visual concepts to downstream tasks. However, current methods explore the fixed adaption of learnable prompt on seen domains, which makes them over-emphasize the primary visual features observed during training, limiting their generalization capabilities to unseen domains. In this work, we propose AENet, which endows semantic information into the visual prompt to distill semantic-enhanced prompt for visual representation enrichment, enabling effective knowledge transfer for ZSL. AENet comprises two key steps: 1) exploring the concept-harmonized tokens for the visual and attribute modalities, grounded on the modal-sharing token that represents consistent visual-semantic concepts; and 2) yielding semantic-enhanced prompt via the visual residual refinement unit with attribute consistency supervision. These are further integrated with primary visual features to attend to semantic-related information for visual enhancement, thus strengthening transferable ability. Experimental results on three benchmarks show that our AENet outperforms existing state-of-the-art ZSL methods. The code is provided in the zip file of supplementary materials.

Lingzhi He, Hongguang Zhu, Feng Li et al.

Depth maps obtained by commercial depth sensors are always in low-resolution, making it difficult to be used in various computer vision tasks. Thus, depth map super-resolution (SR) is a practical and valuable task, which upscales the depth map into high-resolution (HR) space. However, limited by the lack of real-world paired low-resolution (LR) and HR depth maps, most existing methods use downsampling to obtain paired training samples. To this end, we first construct a large-scale dataset named "RGB-D-D", which can greatly promote the study of depth map SR and even more depth-related real-world tasks. The "D-D" in our dataset represents the paired LR and HR depth maps captured from mobile phone and Lucid Helios respectively ranging from indoor scenes to challenging outdoor scenes. Besides, we provide a fast depth map super-resolution (FDSR) baseline, in which the high-frequency component adaptively decomposed from RGB image to guide the depth map SR. Extensive experiments on existing public datasets demonstrate the effectiveness and efficiency of our network compared with the state-of-the-art methods. Moreover, for the real-world LR depth maps, our algorithm can produce more accurate HR depth maps with clearer boundaries and to some extent correct the depth value errors.

Shangrong Yang, Chunyu Lin, Kang Liao et al.

Distortion rectification is often required for fisheye images. The generation-based method is one mainstream solution due to its label-free property, but its naive skip-connection and overburdened decoder will cause blur and incomplete correction. First, the skip-connection directly transfers the image features, which may introduce distortion and cause incomplete correction. Second, the decoder is overburdened during simultaneously reconstructing the content and structure of the image, resulting in vague performance. To solve these two problems, in this paper, we focus on the interpretable correction mechanism of the distortion rectification network and propose a feature-level correction scheme. We embed a correction layer in skip-connection and leverage the appearance flows in different layers to pre-correct the image features. Consequently, the decoder can easily reconstruct a plausible result with the remaining distortion-less information. In addition, we propose a parallel complementary structure. It effectively reduces the burden of the decoder by separating content reconstruction and structure correction. Subjective and objective experiment results on different datasets demonstrate the superiority of our method.

Chongyi Li, Runmin Cong, Chunle Guo et al.

The diverse spatial resolutions, various object types, scales and orientations, and cluttered backgrounds in optical remote sensing images (RSIs) challenge the current salient object detection (SOD) approaches. It is commonly unsatisfactory to directly employ the SOD approaches designed for nature scene images (NSIs) to RSIs. In this paper, we propose a novel Parallel Down-up Fusion network (PDF-Net) for SOD in optical RSIs, which takes full advantage of the in-path low- and high-level features and cross-path multi-resolution features to distinguish diversely scaled salient objects and suppress the cluttered backgrounds. To be specific, keeping a key observation that the salient objects still are salient no matter the resolutions of images are in mind, the PDF-Net takes successive down-sampling to form five parallel paths and perceive scaled salient objects that are commonly existed in optical RSIs. Meanwhile, we adopt the dense connections to take advantage of both low- and high-level information in the same path and build up the relations of cross paths, which explicitly yield strong feature representations. At last, we fuse the multiple-resolution features in parallel paths to combine the benefits of the features with different resolutions, i.e., the high-resolution feature consisting of complete structure and clear details while the low-resolution features highlighting the scaled salient objects. Extensive experiments on the ORSSD dataset demonstrate that the proposed network is superior to the state-of-the-art approaches both qualitatively and quantitatively.

Congqi Cao, Yifan Zhang, Chunjie Zhang et al.

Three dimensional convolutional neural networks (3D CNNs) have been established as a powerful tool to simultaneously learn features from both spatial and temporal dimensions, which is suitable to be applied to video-based action recognition. In this work, we propose not to directly use the activations of fully-connected layers of a 3D CNN as the video feature, but to use selective convolutional layer activations to form a discriminative descriptor for video. It pools the feature on the convolutional layers under the guidance of body joint positions. Two schemes of mapping body joints into convolutional feature maps for pooling are discussed. The body joint positions can be obtained from any off-the-shelf skeleton estimation algorithm. The helpfulness of the body joint guided feature pooling with inaccurate skeleton estimation is systematically evaluated. To make it end-to-end and do not rely on any sophisticated body joint detection algorithm, we further propose a two-stream bilinear model which can learn the guidance from the body joints and capture the spatio-temporal features simultaneously. In this model, the body joint guided feature pooling is conveniently formulated as a bilinear product operation. Experimental results on three real-world datasets demonstrate the effectiveness of body joint guided pooling which achieves promising performance.