Zeyu Chen

Juncai Peng, Yi Liu, Shiyu Tang et al.

Real-world applications have high demands for semantic segmentation methods. Although semantic segmentation has made remarkable leap-forwards with deep learning, the performance of real-time methods is not satisfactory. In this work, we propose PP-LiteSeg, a novel lightweight model for the real-time semantic segmentation task. Specifically, we present a Flexible and Lightweight Decoder (FLD) to reduce computation overhead of previous decoder. To strengthen feature representations, we propose a Unified Attention Fusion Module (UAFM), which takes advantage of spatial and channel attention to produce a weight and then fuses the input features with the weight. Moreover, a Simple Pyramid Pooling Module (SPPM) is proposed to aggregate global context with low computation cost. Extensive evaluations demonstrate that PP-LiteSeg achieves a superior trade-off between accuracy and speed compared to other methods. On the Cityscapes test set, PP-LiteSeg achieves 72.0% mIoU/273.6 FPS and 77.5% mIoU/102.6 FPS on NVIDIA GTX 1080Ti. Source code and models are available at PaddleSeg: https://github.com/PaddlePaddle/PaddleSeg.

Jian Wang, Xiang Long, Guowei Chen et al.

High resolution and advanced semantic representation are both vital for dense prediction. Empirically, low-resolution feature maps often achieve stronger semantic representation, and high-resolution feature maps generally can better identify local features such as edges, but contains weaker semantic information. Existing state-of-the-art frameworks such as HRNet has kept low-resolution and high-resolution feature maps in parallel, and repeatedly exchange the information across different resolutions. However, we believe that the lowest-resolution feature map often contains the strongest semantic information, and it is necessary to go through more layers to merge with high-resolution feature maps, while for high-resolution feature maps, the computational cost of each convolutional layer is very large, and there is no need to go through so many layers. Therefore, we designed a U-shaped High-Resolution Network (U-HRNet), which adds more stages after the feature map with strongest semantic representation and relaxes the constraint in HRNet that all resolutions need to be calculated parallel for a newly added stage. More calculations are allocated to low-resolution feature maps, which significantly improves the overall semantic representation. U-HRNet is a substitute for the HRNet backbone and can achieve significant improvement on multiple semantic segmentation and depth prediction datasets, under the exactly same training and inference setting, with almost no increasing in the amount of calculation. Code is available at PaddleSeg: https://github.com/PaddlePaddle/PaddleSeg.

Guowei Chen, Yi Liu, Jian Wang et al.

Natural image matting is a fundamental and challenging computer vision task. It has many applications in image editing and composition. Recently, deep learning-based approaches have achieved great improvements in image matting. However, most of them require a user-supplied trimap as an auxiliary input, which limits the matting applications in the real world. Although some trimap-free approaches have been proposed, the matting quality is still unsatisfactory compared to trimap-based ones. Without the trimap guidance, the matting models suffer from foreground-background ambiguity easily, and also generate blurry details in the transition area. In this work, we propose PP-Matting, a trimap-free architecture that can achieve high-accuracy natural image matting. Our method applies a high-resolution detail branch (HRDB) that extracts fine-grained details of the foreground with keeping feature resolution unchanged. Also, we propose a semantic context branch (SCB) that adopts a semantic segmentation subtask. It prevents the detail prediction from local ambiguity caused by semantic context missing. In addition, we conduct extensive experiments on two well-known benchmarks: Composition-1k and Distinctions-646. The results demonstrate the superiority of PP-Matting over previous methods. Furthermore, we provide a qualitative evaluation of our method on human matting which shows its outstanding performance in the practical application. The code and pre-trained models will be available at PaddleSeg: https://github.com/PaddlePaddle/PaddleSeg.

Xuhong Li, Haoyi Xiong, Yi Liu et al.

While fine-tuning pre-trained networks has become a popular way to train image segmentation models, such backbone networks for image segmentation are frequently pre-trained using image classification source datasets, e.g., ImageNet. Though image classification datasets could provide the backbone networks with rich visual features and discriminative ability, they are incapable of fully pre-training the target model (i.e., backbone+segmentation modules) in an end-to-end manner. The segmentation modules are left to random initialization in the fine-tuning process due to the lack of segmentation labels in classification datasets. In our work, we propose a method that leverages Pseudo Semantic Segmentation Labels (PSSL), to enable the end-to-end pre-training for image segmentation models based on classification datasets. PSSL was inspired by the observation that the explanation results of classification models, obtained through explanation algorithms such as CAM, SmoothGrad and LIME, would be close to the pixel clusters of visual objects. Specifically, PSSL is obtained for each image by interpreting the classification results and aggregating an ensemble of explanations queried from multiple classifiers to lower the bias caused by single models. With PSSL for every image of ImageNet, the proposed method leverages a weighted segmentation learning procedure to pre-train the segmentation network en masse. Experiment results show that, with ImageNet accompanied by PSSL as the source dataset, the proposed end-to-end pre-training strategy successfully boosts the performance of various segmentation models, i.e., PSPNet-ResNet50, DeepLabV3-ResNet50, and OCRNet-HRNetW18, on a number of segmentation tasks, such as CamVid, VOC-A, VOC-C, ADE20K, and CityScapes, with significant improvements. The source code is availabel at https://github.com/PaddlePaddle/PaddleSeg.

Weibin Liao, Haoyi Xiong, Qingzhong Wang et al. · harvard

While self-supervised learning (SSL) algorithms have been widely used to pre-train deep models, few efforts [11] have been done to improve representation learning of X-ray image analysis with SSL pre-trained models. In this work, we study a novel self-supervised pre-training pipeline, namely Multi-task Self-super-vised Continual Learning (MUSCLE), for multiple medical imaging tasks, such as classification and segmentation, using X-ray images collected from multiple body parts, including heads, lungs, and bones. Specifically, MUSCLE aggregates X-rays collected from multiple body parts for MoCo-based representation learning, and adopts a well-designed continual learning (CL) procedure to further pre-train the backbone subject various X-ray analysis tasks jointly. Certain strategies for image pre-processing, learning schedules, and regularization have been used to solve data heterogeneity, overfitting, and catastrophic forgetting problems for multi-task/dataset learning in MUSCLE.We evaluate MUSCLE using 9 real-world X-ray datasets with various tasks, including pneumonia classification, skeletal abnormality classification, lung segmentation, and tuberculosis (TB) detection. Comparisons against other pre-trained models [7] confirm the proof-of-concept that self-supervised multi-task/dataset continual pre-training could boost the performance of X-ray image analysis.

Jingyi Tang, Gu Wang, Zeyu Chen et al. · tsinghua

Although methods for estimating the pose of objects in indoor scenes have achieved great success, the pose estimation of underwater objects remains challenging due to difficulties brought by the complex underwater environment, such as degraded illumination, blurring, and the substantial cost of obtaining real annotations. In response, we introduce FAFA, a Frequency-Aware Flow-Aided self-supervised framework for 6D pose estimation of unmanned underwater vehicles (UUVs). Essentially, we first train a frequency-aware flow-based pose estimator on synthetic data, where an FFT-based augmentation approach is proposed to facilitate the network in capturing domain-invariant features and target domain styles from a frequency perspective. Further, we perform self-supervised training by enforcing flow-aided multi-level consistencies to adapt it to the real-world underwater environment. Our framework relies solely on the 3D model and RGB images, alleviating the need for any real pose annotations or other-modality data like depths. We evaluate the effectiveness of FAFA on common underwater object pose benchmarks and showcase significant performance improvements compared to state-of-the-art methods. Code is available at github.com/tjy0703/FAFA.

Zeyu Chen, Guanghao Zhou, Qixiang Yin et al.

In this work, we present Valley3, an omni multimodal large language model (MLLM) developed for diverse global e-commerce tasks, with unified understanding and reasoning capabilities across text, images, video, and audio. A key feature of Valley3 is its native multilingual audio capability for e-commerce, developed by extending vision-language models to better support crucial audio-visual tasks, particularly in short-video scenarios. To achieve this, we carefully design a four-stage omni e-commerce continued pre-training pipeline, through which Valley3 progressively acquires audio understanding, cross-modal instruction-following, e-commerce domain knowledge, and long-context reasoning capabilities, ultimately evolving into an omni model for diverse e-commerce scenarios. Then, we further improve Valley3 through post-training to encourage long-chain reasoning with controllable reasoning modes, enabling one non-thinking mode and three distinct levels of thinking, thereby balancing inference efficiency in simple scenarios with deep reasoning for complex applications. Moreover, we equip Valley3 with agentic search capabilities to proactively invoke search tools and acquire task-relevant information for e-commerce deep research tasks. To comprehensively assess the capabilities of Valley3, we construct an omni e-commerce benchmark spanning 6 tasks. Experimental results show that Valley3 consistently outperforms strong baselines on our in-house and open-source e-commerce benchmarks, while remaining competitive on general-domain benchmarks.

Xiaoyu Tang, Zeyu Chen, Jintao Cheng et al.

In autonomous driving, accurately distinguishing between static and moving objects is crucial for the autonomous driving system. When performing the motion object segmentation (MOS) task, effectively leveraging motion information from objects becomes a primary challenge in improving the recognition of moving objects. Previous methods either utilized range view (RV) or bird's eye view (BEV) residual maps to capture motion information. Unlike traditional approaches, we propose combining RV and BEV residual maps to exploit a greater potential of motion information jointly. Thus, we introduce CV-MOS, a cross-view model for moving object segmentation. Novelty, we decouple spatial-temporal information by capturing the motion from BEV and RV residual maps and generating semantic features from range images, which are used as moving object guidance for the motion branch. Our direct and unique solution maximizes the use of range images and RV and BEV residual maps, significantly enhancing the performance of LiDAR-based MOS task. Our method achieved leading IoU(\%) scores of 77.5\% and 79.2\% on the validation and test sets of the SemanticKitti dataset. In particular, CV-MOS demonstrates SOTA performance to date on various datasets. The CV-MOS implementation is available at https://github.com/SCNU-RISLAB/CV-MOS

Yuying Hao, Yi Liu, Yizhou Chen et al.

In recent years, the rapid development of deep learning has brought great advancements to image and video segmentation methods based on neural networks. However, to unleash the full potential of such models, large numbers of high-quality annotated images are necessary for model training. Currently, many widely used open-source image segmentation software relies heavily on manual annotation which is tedious and time-consuming. In this work, we introduce EISeg, an Efficient Interactive SEGmentation annotation tool that can drastically improve image segmentation annotation efficiency, generating highly accurate segmentation masks with only a few clicks. We also provide various domain-specific models for remote sensing, medical imaging, industrial quality inspections, human segmentation, and temporal aware models for video segmentation. The source code for our algorithm and user interface are available at: https://github.com/PaddlePaddle/PaddleSeg.

Xiaoran Fan, Chao Pang, Tian Yuan et al. · apple-ml

Speech representation learning has improved both speech understanding and speech synthesis tasks for single language. However, its ability in cross-lingual scenarios has not been explored. In this paper, we extend the pretraining method for cross-lingual multi-speaker speech synthesis tasks, including cross-lingual multi-speaker voice cloning and cross-lingual multi-speaker speech editing. We propose a speech-text joint pretraining framework, where we randomly mask the spectrogram and the phonemes given a speech example and its transcription. By learning to reconstruct the masked parts of the input in different languages, our model shows great improvements over speaker-embedding-based multi-speaker TTS methods. Moreover, our framework is end-to-end for both the training and the inference without any finetuning effort. In cross-lingual multi-speaker voice cloning and cross-lingual multi-speaker speech editing tasks, our experiments show that our model outperforms speaker-embedding-based multi-speaker TTS methods.

Huijuan Wang, Siming Dai, Weiyue Su et al.

Relational graph neural networks have garnered particular attention to encode graph context in knowledge graphs (KGs). Although they achieved competitive performance on small KGs, how to efficiently and effectively utilize graph context for large KGs remains an open problem. To this end, we propose the Relation-based Embedding Propagation (REP) method. It is a post-processing technique to adapt pre-trained KG embeddings with graph context. As relations in KGs are directional, we model the incoming head context and the outgoing tail context separately. Accordingly, we design relational context functions with no external parameters. Besides, we use averaging to aggregate context information, making REP more computation-efficient. We theoretically prove that such designs can avoid information distortion during propagation. Extensive experiments also demonstrate that REP has significant scalability while improving or maintaining prediction quality. Notably, it averagely brings about 10% relative improvement to triplet-based embedding methods on OGBL-WikiKG2 and takes 5%-83% time to achieve comparable results as the state-of-the-art GC-OTE.

Yuying Hao, Yi Liu, Juncai Peng et al.

Interactive image segmentation aims at segmenting a target region through a way of human-computer interaction. Recent works based on deep learning have achieved excellent performance, while most of them focus on improving the accuracy of the training set and ignore potential improvement on the test set. In the inference phase, they tend to have a good performance on similar domains to the training set, and lack adaptability to domain shift, so they require more user efforts to obtain satisfactory results. In this work, we propose RAIS, a robust and accurate architecture for interactive segmentation with continuous learning, where the model can learn from both train and test data sets. For efficient learning on the test set, we propose a novel optimization strategy to update global and local parameters with a basic segmentation module and adaptation module, respectively. Moreover, we perform extensive experiments on several benchmarks that show our method can handle data distribution shifts and achieves SOTA performance compared with recent interactive segmentation methods. Besides, our method also shows its robustness in the datasets of remote sensing and medical imaging where the data domains are completely different between training and testing.

Haifeng Wang, Hua Wu, Tian Wu et al.

In this report, we introduce ERNIE 5.0, a natively autoregressive foundation model desinged for unified multimodal understanding and generation across text, image, video, and audio. All modalities are trained from scratch under a unified next-group-of-tokens prediction objective, based on an ultra-sparse mixture-of-experts (MoE) architecture with modality-agnostic expert routing. To address practical challenges in large-scale deployment under diverse resource constraints, ERNIE 5.0 adopts a novel elastic training paradigm. Within a single pre-training run, the model learns a family of sub-models with varying depths, expert capacities, and routing sparsity, enabling flexible trade-offs among performance, model size, and inference latency in memory- or time-constrained scenarios. Moreover, we systematically address the challenges of scaling reinforcement learning to unified foundation models, thereby guaranteeing efficient and stable post-training under ultra-sparse MoE architectures and diverse multimodal settings. Extensive experiments demonstrate that ERNIE 5.0 achieves strong and balanced performance across multiple modalities. To the best of our knowledge, among publicly disclosed models, ERNIE 5.0 represents the first production-scale realization of a trillion-parameter unified autoregressive model that supports both multimodal understanding and generation. To facilitate further research, we present detailed visualizations of modality-agnostic expert routing in the unified model, alongside comprehensive empirical analysis of elastic training, aiming to offer profound insights to the community.

Hanzhuo Tan, Chunpu Xu, Jing Li et al.

Natural language understanding (NLU) is integral to various social media applications. However, existing NLU models rely heavily on context for semantic learning, resulting in compromised performance when faced with short and noisy social media content. To address this issue, we leverage in-context learning (ICL), wherein language models learn to make inferences by conditioning on a handful of demonstrations to enrich the context and propose a novel hashtag-driven in-context learning (HICL) framework. Concretely, we pre-train a model #Encoder, which employs #hashtags (user-annotated topic labels) to drive BERT-based pre-training through contrastive learning. Our objective here is to enable #Encoder to gain the ability to incorporate topic-related semantic information, which allows it to retrieve topic-related posts to enrich contexts and enhance social media NLU with noisy contexts. To further integrate the retrieved context with the source text, we employ a gradient-based method to identify trigger terms useful in fusing information from both sources. For empirical studies, we collected 45M tweets to set up an in-context NLU benchmark, and the experimental results on seven downstream tasks show that HICL substantially advances the previous state-of-the-art results. Furthermore, we conducted extensive analyzes and found that: (1) combining source input with a top-retrieved post from #Encoder is more effective than using semantically similar posts; (2) trigger words can largely benefit in merging context from the source and retrieved posts.

Zeyu Chen, Jie Li, Kai Han

Multimodal representation alignment is pivotal for large language models and robotics. Traditional methods are often hindered by cross-modal information discrepancies and data scarcity, leading to suboptimal alignment spaces that overlook modality-unique features. We propose CodeBind, a framework that optimizes multimodal representation spaces through a modality-shared-specific codebook design. By incrementally aligning target and bridging modalities, CodeBind bypasses the need for fully paired data. Unlike traditional hard alignment, CodeBind decomposes features into shared components for semantic consistency and specific components for modality-unique details. This design utilizes a compositional vector quantization scheme, where a shared codebook bridges modality gaps and modality-specific codebooks mitigate representation bias by preventing dominant modalities from overshadowing others. Validated across nine modalities (text, image, video, audio, depth, thermal, tactile, 3D point cloud, EEG), CodeBind achieves state-of-the-art performance in multimodal classification and retrieval tasks.

Yan Shu, Weichao Zeng, Fangmin Zhao et al.

Visual text is a crucial component in both document and scene images, conveying rich semantic information and attracting significant attention in the computer vision community. Beyond traditional tasks such as text detection and recognition, visual text processing has witnessed rapid advancements driven by the emergence of foundation models, including text image reconstruction and text image manipulation. Despite significant progress, challenges remain due to the unique properties that differentiate text from general objects. Effectively capturing and leveraging these distinct textual characteristics is essential for developing robust visual text processing models. In this survey, we present a comprehensive, multi-perspective analysis of recent advancements in visual text processing, focusing on two key questions: (1) What textual features are most suitable for different visual text processing tasks? (2) How can these distinctive text features be effectively incorporated into processing frameworks? Furthermore, we introduce VTPBench, a new benchmark that encompasses a broad range of visual text processing datasets. Leveraging the advanced visual quality assessment capabilities of multimodal large language models (MLLMs), we propose VTPScore, a novel evaluation metric designed to ensure fair and reliable evaluation. Our empirical study with more than 20 specific models reveals substantial room for improvement in the current techniques. Our aim is to establish this work as a fundamental resource that fosters future exploration and innovation in the dynamic field of visual text processing. The relevant repository is available at https://github.com/shuyansy/Visual-Text-Processing-survey.

Chenmin Yu, Liu Yu, Daiqing Wu et al.

Modern visual object trackers show impressive results on general targets, yet their performance drops substantially when dealing with scene text. Although currently underexplored, tracking text in videos is essential for dynamic text manipulations such as segmentation, removal, and editing. To fill this gap, this paper formalizes this specific task as Scene Text Tracking and presents the first systematic work for it. We identify three primary challenges in this task: 1) severe geometric distortions from perspective shifts, 2) high visual ambiguity across different instances, and 3) high sensitivity to fine-grained structural details. To address these issues, we propose SymTrack, a unified detection-free framework with synergistic dual-branch design. It integrates a Cross-Expert Calibration mechanism to reduce semantic bias, along with a Predictive Token Rectification mechanism to correct structural imbalances, complemented by an Adaptive Inference Engine that stabilizes predictions under motion constraints. Considering the lack of dedicated benchmarks for this task, we utilize three datasets from video text spotting to construct a benchmark with high-quality annotations. Extensive experiments demonstrate that SymTrack sets the new state-of-the-art on all three benchmarks, outperforming previous best trackers by up to 11.97\% AUC on $ \text{BOVText}_{\text{SOT}} $. Overall, our work promotes efficient and thorough text tracking, paving the way toward more generalized video text manipulation.

Yichao Liu, Huawen Shen, Liu Yu et al.

GUI agents powered by Multimodal Large Language Models (MLLMs) have demonstrated impressive capability in understanding and executing user instructions. However, accurately grounding instruction-relevant elements from high-resolution screenshots cluttered with irrelevant UI components remains challenging for existing approaches. Inspired by how humans dynamically adjust their perceptual scope to locate task-related regions on complex screens, we propose DRS-GUI, a training-free dynamic region search framework for GUI grounding that can be seamlessly integrated into existing MLLMs. DRS-GUI introduces a lightweight UI Perceptor that performs three human-like perceptual actions (Focus, Shift, and Scatter) to progressively explore the interface and generate region proposals. To dynamically schedule these actions, we further design an Action Planner based on Monte Carlo Tree Search (MCTS). A region quality reward is employed to evaluate and select the highly instruction-relevant region, efficiently pruning redundant UI elements. Experiments demonstrate that DRS-GUI yields a 14\% improvement on ScreenSpot-Pro for general and GUI-specific MLLMs (Qwen2.5-VL-7B and UGround-V1-7B), significantly enhancing grounding performance and generalization.

Xu Zhang, Hao Zhou, Haoming Qin et al.

Despite substantial progress in text-to-video generation, achieving precise and flexible control over fine-grained spatiotemporal attributes remains a significant unresolved challenge in video generation research. To address these limitations, we introduce VCtrl (also termed PP-VCtrl), a novel framework designed to enable fine-grained control over pre-trained video diffusion models in a unified manner. VCtrl integrates diverse user-specified control signals-such as Canny edges, segmentation masks, and human keypoints-into pretrained video diffusion models via a generalizable conditional module capable of uniformly encoding multiple types of auxiliary signals without modifying the underlying generator. Additionally, we design a unified control signal encoding pipeline and a sparse residual connection mechanism to efficiently incorporate control representations. Comprehensive experiments and human evaluations demonstrate that VCtrl effectively enhances controllability and generation quality. The source code and pre-trained models are publicly available and implemented using the PaddlePaddle framework at http://github.com/PaddlePaddle/PaddleMIX/tree/develop/ppdiffusers/examples/ppvctrl.

Tianyi Chen, Tianyu Ding, Zhihui Zhu et al.

Compressing a predefined deep neural network (DNN) into a compact sub-network with competitive performance is crucial in the efficient machine learning realm. This topic spans various techniques, from structured pruning to neural architecture search, encompassing both pruning and erasing operators perspectives. Despite advancements, existing methods suffers from complex, multi-stage processes that demand substantial engineering and domain knowledge, limiting their broader applications. We introduce the third-generation Only-Train-Once (OTOv3), which first automatically trains and compresses a general DNN through pruning and erasing operations, creating a compact and competitive sub-network without the need of fine-tuning. OTOv3 simplifies and automates the training and compression process, minimizes the engineering efforts required from users. It offers key technological advancements: (i) automatic search space construction for general DNNs based on dependency graph analysis; (ii) Dual Half-Space Projected Gradient (DHSPG) and its enhanced version with hierarchical search (H2SPG) to reliably solve (hierarchical) structured sparsity problems and ensure sub-network validity; and (iii) automated sub-network construction using solutions from DHSPG/H2SPG and dependency graphs. Our empirical results demonstrate the efficacy of OTOv3 across various benchmarks in structured pruning and neural architecture search. OTOv3 produces sub-networks that match or exceed the state-of-the-arts. The source code will be available at https://github.com/tianyic/only_train_once.

Dongjing Shan, Yamei Luo, Jiqing Xuan et al.

Early detection of myometrial invasion is critical for the staging and life-saving management of endometrial carcinoma (EC), a prevalent global malignancy. Transvaginal ultrasound serves as the primary, accessible screening modality in resource-constrained primary care settings; however, its diagnostic reliability is severely hindered by low tissue contrast, high operator dependence, and a pronounced scarcity of positive pathological samples. Existing artificial intelligence solutions struggle to overcome this severe class imbalance and the subtle imaging features of invasion, particularly under the strict computational limits of primary care clinics. Here we present an automated, highly efficient two-stage deep learning framework that resolves both data and computational bottlenecks in EC screening. To mitigate pathological data scarcity, we develop a structure-guided cross-modal generation network that synthesizes diverse, high-fidelity ultrasound images from unpaired magnetic resonance imaging (MRI) data, strictly preserving clinically essential anatomical junctions. Furthermore, we introduce a lightweight screening network utilizing gradient distillation, which transfers discriminative knowledge from a high-capacity teacher model to dynamically guide sparse attention towards task-critical regions. Evaluated on a large, multicenter cohort of 7,951 participants, our model achieves a sensitivity of 99.5\%, a specificity of 97.2\%, and an area under the curve of 0.987 at a minimal computational cost (0.289 GFLOPs), substantially outperforming the average diagnostic accuracy of expert sonographers. Our approach demonstrates that combining cross-modal synthetic augmentation with knowledge-driven efficient modeling can democratize expert-level, real-time cancer screening for resource-constrained primary care settings.

Feng Ni, Kui Huang, Yao Lu et al.

With the rapid advancement of digitalization, various document images are being applied more extensively in production and daily life, and there is an increasingly urgent need for fast and accurate parsing of the content in document images. Therefore, this report presents PP-DocBee, a novel multimodal large language model designed for end-to-end document image understanding. First, we develop a data synthesis strategy tailored to document scenarios in which we build a diverse dataset to improve the model generalization. Then, we apply a few training techniques, including dynamic proportional sampling, data preprocessing, and OCR postprocessing strategies. Extensive evaluations demonstrate the superior performance of PP-DocBee, achieving state-of-the-art results on English document understanding benchmarks and even outperforming existing open source and commercial models in Chinese document understanding. The source code and pre-trained models are publicly available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.

Zeyu Chen, Fangmin Zhao, Yan Shu et al.

Style-conditioned scene text generation faces unique challenges in extracting precise text styles from complex backgrounds and maintaining fine-grained style consistency across characters, especially for multilingual scripts. We propose StyleTextGen, a novel framework that learns to perceive and replicate visual text styles across different languages and writing systems. Our approach features three key contributions: First, we introduce a dual-branch style encoder dedicated to style modeling, yielding robust multilingual text style representations in complex real-world scenes. Second, we design a text style consistency loss that enhances style coherence and improves overall visual quality. Third, we develop a mask-guided inference strategy that ensures precise style alignment between generated and reference text. To facilitate systematic evaluation, we construct StyleText-CE, a bilingual scene text style benchmark covering both monolingual and cross-lingual settings. Extensive experiments demonstrate that StyleTextGen significantly outperforms existing methods in style consistency and cross-lingual generalization, establishing new state-of-the-art performance in multilingual style-conditioned text generation.

Shihan Chen, Zhaojin Li, Zeyu Chen et al.

Recent developments in 3D Gaussian Splatting have made significant advances in surface reconstruction. However, scaling these methods to large-scale scenes remains challenging due to high computational demands and the complex dynamic appearances typical of outdoor environments. These challenges hinder the application in aerial surveying and autonomous driving. This paper proposes a novel solution to reconstruct large-scale surfaces with fine details, supervised by full-sized images. Firstly, we introduce a coarse-to-fine strategy to reconstruct a coarse model efficiently, followed by adaptive scene partitioning and sub-scene refining from image segments. Additionally, we integrate a decoupling appearance model to capture global appearance variations and a transient mask model to mitigate interference from moving objects. Finally, we expand the multi-view constraint and introduce a single-view regularization for texture-less areas. Our experiments were conducted on the publicly available dataset GauU-Scene V2, which was captured using unmanned aerial vehicles. To the best of our knowledge, our method outperforms existing NeRF-based and Gaussian-based methods, achieving high-fidelity visual results and accurate surface from full-size image optimization. Open-source code will be available on GitHub.

Wenting Li, Jiahong Xue, Xi Zhang et al.

Word embedding has become ubiquitous and is widely used in various natural language processing (NLP) tasks, such as web retrieval, web semantic analysis, and machine translation, and so on. Unfortunately, training the word embedding in a relatively large corpus is prohibitively expensive. We propose a graph-based word embedding algorithm, called Word-Graph2vec, which converts the large corpus into a word co-occurrence graph, then takes the word sequence samples from this graph by randomly traveling and trains the word embedding on this sampling corpus in the end. We posit that because of the limited vocabulary, huge idioms, and fixed expressions in English, the size and density of the word co-occurrence graph change slightly with the increase in the training corpus. So that Word-Graph2vec has stable runtime on the large-scale data set, and its performance advantage becomes more and more obvious with the growth of the training corpus. Extensive experiments conducted on real-world datasets show that the proposed algorithm outperforms traditional Word2vec four to five times in terms of efficiency and two to three times than FastText, while the error generated by the random walk technique is small.

Zeyu Chen, Junde Wu

This paper identifies exact probabilistic simulation cost as the natural quantitative measure of quantum advantage for finite automata under strict cutpoints. It gives sharp simulation laws for two representative models. A one-way finite automaton with $c$ classical states and a $q$-dimensional quantum register has exact probabilistic simulation cost $Θ(cq^2)$, while an $n$-dimensional measure-once one-way quantum finite automaton has worst-case cost $Θ(n^2)$. The proofs develop a prepare--test framework, in which prefixes generate the relevant real operator degrees of freedom and suffixes convert them into strict-cutpoint tests. The same obstruction is recast through finite sign-rank matrices, clarifying the role of Forster's spectral method. Placed beside the surrounding two-way separations, these results give a clean hierarchy of finite-automata quantum advantage.

Yifu Luo, Xinhao Hu, Keyu Fan et al.

Reinforcement learning (RL) has garnered increasing attention in text-to-image (T2I) generation. However, most existing RL approaches are tailored to either diffusion models or autoregressive models, overlooking an important alternative: masked generative models. In this work, we propose Mask-GRPO, the first method to incorporate Group Relative Policy Optimization (GRPO)-based RL into this overlooked paradigm. Our core insight is to redefine the transition probability, which is different from current approaches, and formulate the unmasking process as a multi-step decision-making problem. To further enhance our method, we explore several useful strategies, including removing the KL constraint, applying the reduction strategy, and filtering out low-quality samples. Using Mask-GRPO, we improve a base model, Show-o, with substantial improvements on standard T2I benchmarks and preference alignment, outperforming existing state-of-the-art approaches. The code is available on https://github.com/xingzhejun/Mask-GRPO

Zeyu Chen, Junde Wu

Affine automata provide a finite-state computational model that preserves the linear-algebraic structure of quantum computation while operating entirely over the reals. Recent work has shown that affine automata can far surpass classical probabilistic finite-state verifiers. However, prior constructions relied on arbitrary real-valued transition matrices, leaving open whether the observed power stems from the affine mechanism itself or from computational resources implicitly encoded in irrational or infinite-precision parameters. This paper studies one-way and two-way automata with deterministic and affine states as verifiers in Arthur--Merlin proof systems under the restriction that every affine transition matrix has rational entries, and shows that the resulting rational model still supports the main verification advantages of affine finite-state verification. At the one-way level, we verify benchmark nonregular languages that are provably hard or impossible for classical two-way probabilistic verifiers. At the two-way level, we achieve weak verification of every Turing-recognizable language, strong bounded-error verification for every language in $\mathbf{ATIME}(2^{O(n)})$, and perfect-completeness strong verification for every language in $\mathbf{PSPACE}$. These results establish that the remarkable verification power of affine finite-state automata is structural.

Chunyu Cao, Jintao Cheng, Zeyu Chen et al.

Motion Object Segmentation (MOS) is crucial for autonomous driving, as it enhances localization, path planning, map construction, scene flow estimation, and future state prediction. While existing methods achieve strong performance, balancing accuracy and real-time inference remains a challenge. To address this, we propose a logits-based knowledge distillation framework for MOS, aiming to improve accuracy while maintaining real-time efficiency. Specifically, we adopt a Bird's Eye View (BEV) projection-based model as the student and a non-projection model as the teacher. To handle the severe imbalance between moving and non-moving classes, we decouple them and apply tailored distillation strategies, allowing the teacher model to better learn key motion-related features. This approach significantly reduces false positives and false negatives. Additionally, we introduce dynamic upsampling, optimize the network architecture, and achieve a 7.69% reduction in parameter count, mitigating overfitting. Our method achieves a notable IoU of 78.8% on the hidden test set of the SemanticKITTI-MOS dataset and delivers competitive results on the Apollo dataset. The KDMOS implementation is available at https://github.com/SCNU-RISLAB/KDMOS.

Lutao Chu, Yi Liu, Zewu Wu et al.

As the COVID-19 pandemic rampages across the world, the demands of video conferencing surge. To this end, real-time portrait segmentation becomes a popular feature to replace backgrounds of conferencing participants. While feature-rich datasets, models and algorithms have been offered for segmentation that extract body postures from life scenes, portrait segmentation has yet not been well covered in a video conferencing context. To facilitate the progress in this field, we introduce an open-source solution named PP-HumanSeg. This work is the first to construct a large-scale video portrait dataset that contains 291 videos from 23 conference scenes with 14K fine-labeled frames and extensions to multi-camera teleconferencing. Furthermore, we propose a novel Semantic Connectivity-aware Learning (SCL) for semantic segmentation, which introduces a semantic connectivity-aware loss to improve the quality of segmentation results from the perspective of connectivity. And we propose an ultra-lightweight model with SCL for practical portrait segmentation, which achieves the best trade-off between IoU and the speed of inference. Extensive evaluations on our dataset demonstrate the superiority of SCL and our model. The source code is available at https://github.com/PaddlePaddle/PaddleSeg.

Yuying Hao, Yi Liu, Zewu Wu et al.

High-quality training data play a key role in image segmentation tasks. Usually, pixel-level annotations are expensive, laborious and time-consuming for the large volume of training data. To reduce labelling cost and improve segmentation quality, interactive segmentation methods have been proposed, which provide the result with just a few clicks. However, their performance does not meet the requirements of practical segmentation tasks in terms of speed and accuracy. In this work, we propose EdgeFlow, a novel architecture that fully utilizes interactive information of user clicks with edge-guided flow. Our method achieves state-of-the-art performance without any post-processing or iterative optimization scheme. Comprehensive experiments on benchmarks also demonstrate the superiority of our method. In addition, with the proposed method, we develop an efficient interactive segmentation tool for practical data annotation tasks. The source code and tool is avaliable at https://github.com/PaddlePaddle/PaddleSeg.

Yi Liu, Lutao Chu, Guowei Chen et al.

Image Segmentation plays an essential role in computer vision and image processing with various applications from medical diagnosis to autonomous car driving. A lot of segmentation algorithms have been proposed for addressing specific problems. In recent years, the success of deep learning techniques has tremendously influenced a wide range of computer vision areas, and the modern approaches of image segmentation based on deep learning are becoming prevalent. In this article, we introduce a high-efficient development toolkit for image segmentation, named PaddleSeg. The toolkit aims to help both developers and researchers in the whole process of designing segmentation models, training models, optimizing performance and inference speed, and deploying models. Currently, PaddleSeg supports around 20 popular segmentation models and more than 50 pre-trained models from real-time and high-accuracy levels. With modular components and backbone networks, users can easily build over one hundred models for different requirements. Furthermore, we provide comprehensive benchmarks and evaluations to show that these segmentation algorithms trained on our toolkit have more competitive accuracy. Also, we provide various real industrial applications and practical cases based on PaddleSeg. All codes and examples of PaddleSeg are available at https://github.com/PaddlePaddle/PaddleSeg.

Di Jiang, Zeyu Chen, Rongzhong Lian et al.

Familia is an open-source toolkit for pragmatic topic modeling in industry. Familia abstracts the utilities of topic modeling in industry as two paradigms: semantic representation and semantic matching. Efficient implementations of the two paradigms are made publicly available for the first time. Furthermore, we provide off-the-shelf topic models trained on large-scale industrial corpora, including Latent Dirichlet Allocation (LDA), SentenceLDA and Topical Word Embedding (TWE). We further describe typical applications which are successfully powered by topic modeling, in order to ease the confusions and difficulties of software engineers during topic model selection and utilization.

Zeyu Chen, Pengfei Zhang, Kai Ye et al.

The burgeoning short video industry has accelerated the advancement of video-music retrieval technology, assisting content creators in selecting appropriate music for their videos. In self-supervised training for video-to-music retrieval, the video and music samples in the dataset are separated from the same video work, so they are all one-to-one matches. This does not match the real situation. In reality, a video can use different music as background music, and a music can be used as background music for different videos. Many videos and music that are not in a pair may be compatible, leading to false negative noise in the dataset. A novel inter-intra modal (II) loss is proposed as a solution. By reducing the variation of feature distribution within the two modalities before and after the encoder, II loss can reduce the model's overfitting to such noise without removing it in a costly and laborious way. The video-music retrieval framework, II-CLVM (Contrastive Learning for Video-Music Retrieval), incorporating the II Loss, achieves state-of-the-art performance on the YouTube8M dataset. The framework II-CLVTM shows better performance when retrieving music using multi-modal video information (such as text in videos). Experiments are designed to show that II loss can effectively alleviate the problem of false negative noise in retrieval tasks. Experiments also show that II loss improves various self-supervised and supervised uni-modal and cross-modal retrieval tasks, and can obtain good retrieval models with a small amount of training samples.

Zeyu Chen, Junde Wu

A basic question in the theory of two-state measure-once quantum finite automata (MO-QFAs) is whether two distinct input words can be separated with certainty. In the setting considered here, this exact separation problem reduces to a trace-vanishing question in \(SU(2)\): given distinct positive words \(u\) and \(v\), find matrices \(A,B\in SU(2)\) such that the evaluated trace of \(u^{-1}v\) is zero. The central difficulty lies in the genuinely nonabelian regime where \(u\) and \(v\) have the same abelianization, so the obvious commutative information disappears and the fine structure of the word must be connected to the geometry of representations. This paper develops a slice-driven framework for that task and proves exact separation for every hard positive-word difference covered by four explicit certified conditions, thereby reducing the problem to a sharply delimited residual super-degenerate class. The method extracts algebraic data from the positive-word difference and uses them to select explicit low-dimensional families in \(SU(2)^2\) on which the trace becomes computable. On the algebraic side, the metabelian polynomial is decomposed into explicit interval blocks determined by prefix statistics, and a suitable slope specialization preserves nontrivial information. On the analytic side, the paper derives a computable quadratic trace identity on a visible one-parameter family and complements it with a Laurent-matrix sum-of-squares identity in a parallel algebraic model. These certified criteria are already strong in numerical experiments. This paper also shows that no method based only on finitely many finite-image tests can be universal.

Kunyi Li, Michael Niemeyer, Zeyu Chen et al.

Accurate meshing from monocular images remains a key challenge in 3D vision. While state-of-the-art 3D Gaussian Splatting (3DGS) methods excel at synthesizing photorealistic novel views through rasterization-based rendering, their reliance on sparse, explicit primitives severely limits their ability to recover watertight and topologically consistent 3D surfaces.We introduce MonoGSDF, a novel method that couples Gaussian-based primitives with a neural Signed Distance Field (SDF) for high-quality reconstruction. During training, the SDF guides Gaussians' spatial distribution, while at inference, Gaussians serve as priors to reconstruct surfaces, eliminating the need for memory-intensive Marching Cubes. To handle arbitrary-scale scenes, we propose a scaling strategy for robust generalization. A multi-resolution training scheme further refines details and monocular geometric cues from off-the-shelf estimators enhance reconstruction quality. Experiments on real-world datasets show MonoGSDF outperforms prior methods while maintaining efficiency.

Zeyu Chen, Junde Wu

Generalized finite automata (GFAs), probabilistic finite automata (PFAs), and one-way general quantum finite automata (1gQFA) recognize the same strict-cutpoint languages, but the state complexity of exact probabilistic simulation has remained unclear. This paper determines that worst-case cost exactly: every \(n\)-state 1gQFA admits exact strict-cutpoint simulation by a one-way PFA with \(O(n^2)\) states, via the standard \(n^2\)-dimensional mixed-state linearization together with an explicit alphabet-preserving construction that converts each \(k\)-state GFA into a one-way PFA with at most \(2k+6\) states; conversely, for every \(n\ge 2\), there exists an \(n\)-state 1gQFA for which every equivalent one-way PFA requires at least \(n^2-1\) states, obtained from a prepare--test construction and a Vapnik--Chervonenkis dimension argument. Hence the worst-case probabilistic state cost of exact strict-cutpoint simulation is \(Θ(n^2)\).

Minjia Mao, Dongjun Wei, Zeyu Chen et al.

Recent advancements in large language models (LLMs) have highlighted the risk of misusing them, raising the need for accurate detection of LLM-generated content. In response, a viable solution is to inject imperceptible identifiers into LLMs, known as watermarks. Our research extends the existing watermarking methods by proposing the novel Sampling One Then Accepting (STA-1) method. STA-1 is an unbiased watermark that preserves the original token distribution in expectation and has a lower risk of producing unsatisfactory outputs in low-entropy scenarios compared to existing unbiased watermarks. In watermark detection, STA-1 does not require prompts or a white-box LLM, provides statistical guarantees, demonstrates high efficiency in detection time, and remains robust against various watermarking attacks. Experimental results on low-entropy and high-entropy datasets demonstrate that STA-1 achieves the above properties simultaneously, making it a desirable solution for watermarking LLMs. Implementation codes for this study are available online.

Yun Ji, Zeyu Chen, Xiaoxiong Zhong et al.

Multi-server Federated Learning (FL) has emerged as a promising solution to mitigate communication bottlenecks of single-server FL. We focus on a typical multi-server FL architecture, where the regions covered by different edge servers (ESs) may overlap. A key observation of this architecture is that clients located in the overlapping areas can access edge models from multiple ESs. Building on this insight, we propose FedOC (Federated learning with Overlapping Clients), a novel framework designed to fully exploit the potential of these overlapping clients. In FedOC, overlapping clients could serve dual roles: (1) as Relay Overlapping Clients (ROCs), they forward edge models between neighboring ESs in real time to facilitate model sharing among different ESs; and (2) as Normal Overlapping Clients (NOCs), they dynamically select their initial model for local training based on the edge model delivery time, which enables indirect data fusion among different regions of ESs. The overall FedOC workflow proceeds as follows: in every round, each client trains local model based on the earliest received edge model and transmits to the respective ESs for model aggregation. Then each ES transmits the aggregated edge model to neighboring ESs through ROC relaying. Upon receiving the relayed models, each ES performs a second aggregation and subsequently broadcasts the updated model to covered clients. The existence of ROCs enables the model of each ES to be disseminated to the other ESs in a decentralized manner, which indirectly achieves intercell model and speeding up the training process, making it well-suited for latency-sensitive edge environments. Extensive experimental results show remarkable performance gains of our scheme compared to existing methods.

Bin Wang, Fei Deng, Zeyu Chen et al.

Source-Free Domain Adaptation (SFDA) enables domain adaptation for semantic segmentation of Remote Sensing Images (RSIs) using only a well-trained source model and unlabeled target domain data. However, the lack of ground-truth labels in the target domain often leads to the generation of noisy pseudo-labels. Such noise impedes the effective mitigation of domain shift (DS). To address this challenge, we propose ProSFDA, a prototype-guided SFDA framework. It employs prototype-weighted pseudo-labels to facilitate reliable self-training (ST) under pseudo-labels noise. We, in addition, introduce a prototype-contrast strategy that encourages the aggregation of features belonging to the same class, enabling the model to learn discriminative target domain representations without relying on ground-truth supervision. Extensive experiments show that our approach substantially outperforms existing methods.

Zeyu Chen, Wen Chen, Jun Li et al.

Federated Learning (FL) offers a decentralized solution that allows collaborative local model training and global aggregation, thereby protecting data privacy. In conventional FL frameworks, data privacy is typically preserved under the assumption that local data remains absolutely private, whereas the mobility of clients is frequently neglected in explicit modeling. In this paper, we propose a hierarchical federated learning framework based on the social network with mobility namely HFL-SNM that considers both data sharing among clients and their mobility patterns. Under the constraints of limited resources, we formulate a joint optimization problem of resource allocation and client scheduling, which objective is to minimize the energy consumption of clients during the FL process. In social network, we introduce the concepts of Effective Data Coverage Rate and Redundant Data Coverage Rate. We analyze the impact of effective data and redundant data on the model performance through preliminary experiments. We decouple the optimization problem into multiple sub-problems, analyze them based on preliminary experimental results, and propose Dynamic Optimization in Social Network with Mobility (DO-SNM) algorithm. Experimental results demonstrate that our algorithm achieves superior model performance while significantly reducing energy consumption, compared to traditional baseline algorithms.

Hanqi Chen, Xu Zhang, Xiaoliu Guan et al.

Diffusion Transformers (DiTs) have demonstrated remarkable generative capabilities, particularly benefiting from Transformer architectures that enhance visual and artistic fidelity. However, their inherently sequential denoising process results in high inference latency, limiting their deployment in real-time scenarios. Existing training-free acceleration approaches typically reuse intermediate features at fixed timesteps or layers, overlooking the evolving semantic focus across denoising stages and Transformer blocks.To address this, we propose Sortblock, a training-free inference acceleration framework that dynamically caches block-wise features based on their similarity across adjacent timesteps. By ranking the evolution of residuals, Sortblock adaptively determines a recomputation ratio, selectively skipping redundant computations while preserving generation quality. Furthermore, we incorporate a lightweight linear prediction mechanism to reduce accumulated errors in skipped blocks.Extensive experiments across various tasks and DiT architectures demonstrate that Sortblock achieves over 2$\times$ inference speedup with minimal degradation in output quality, offering an effective and generalizable solution for accelerating diffusion-based generative models.

Weiyue Su, Zeyang Fang, Hui Zhong et al.

WikiKG90M in KDD Cup 2021 is a large encyclopedic knowledge graph, which could benefit various downstream applications such as question answering and recommender systems. Participants are invited to complete the knowledge graph by predicting missing triplets. Recent representation learning methods have achieved great success on standard datasets like FB15k-237. Thus, we train the advanced algorithms in different domains to learn the triplets, including OTE, QuatE, RotatE and TransE. Significantly, we modified OTE into NOTE (short for Norm-OTE) for better performance. Besides, we use both the DeepWalk and the post-smoothing technique to capture the graph structure for supplementation. In addition to the representations, we also use various statistical probabilities among the head entities, the relations and the tail entities for the final prediction. Experimental results show that the ensemble of state-of-the-art representation learning methods could draw on each others strengths. And we develop feature engineering from validation candidates for further improvements. Please note that we apply the same strategy on the test set for final inference. And these features may not be practical in the real world when considering ranking against all the entities.

Zeyu Chen, Xinhang Zhang, Juan Li et al.

Fatigue detection is valued for people to keep mental health and prevent safety accidents. However, detecting facial fatigue, especially mild fatigue in the real world via machine vision is still a challenging issue due to lack of non-lab dataset and well-defined algorithms. In order to improve the detection capability on facial fatigue that can be used widely in daily life, this paper provided an audiovisual dataset named DLFD (daily-life fatigue dataset) which reflected people's facial fatigue state in the wild. A framework using 3D-ResNet along with non-local attention mechanism was training for extraction of local and long-range features in spatial and temporal dimensions. Then, a compacted loss function combining mean squared error and cross-entropy was designed to predict both continuous and categorical fatigue degrees. Our proposed framework has reached an average accuracy of 90.8% on validation set and 72.5% on test set for binary classification, standing a good position compared to other state-of-the-art methods. The analysis of feature map visualization revealed that our framework captured facial dynamics and attempted to build a connection with fatigue state. Our experimental results in multiple metrics proved that our framework captured some typical, micro and dynamic facial features along spatiotemporal dimensions, contributing to the mild fatigue detection in the wild.

Siyu Huang, Haoyi Xiong, Tianyang Wang et al.

Arbitrary image style transfer is a challenging task which aims to stylize a content image conditioned on arbitrary style images. In this task the feature-level content-style transformation plays a vital role for proper fusion of features. Existing feature transformation algorithms often suffer from loss of content or style details, non-natural stroke patterns, and unstable training. To mitigate these issues, this paper proposes a new feature-level style transformation technique, named Style Projection, for parameter-free, fast, and effective content-style transformation. This paper further presents a real-time feed-forward model to leverage Style Projection for arbitrary image style transfer, which includes a regularization term for matching the semantics between input contents and stylized outputs. Extensive qualitative analysis, quantitative evaluation, and user study have demonstrated the effectiveness and efficiency of the proposed methods.

Zhi Fengy, Haoyi Xiong, Chuanyuan Song et al.

Federated machine learning systems have been widely used to facilitate the joint data analytics across the distributed datasets owned by the different parties that do not trust each others. In this paper, we proposed a novel Gradient Boosting Machines (GBM) framework SecureGBM built-up with a multi-party computation model based on semi-homomorphic encryption, where every involved party can jointly obtain a shared Gradient Boosting machines model while protecting their own data from the potential privacy leakage and inferential identification. More specific, our work focused on a specific "dual--party" secure learning scenario based on two parties -- both party own an unique view (i.e., attributes or features) to the sample group of samples while only one party owns the labels. In such scenario, feature and label data are not allowed to share with others. To achieve the above goal, we firstly extent -- LightGBM -- a well known implementation of tree-based GBM through covering its key operations for training and inference with SEAL homomorphic encryption schemes. However, the performance of such re-implementation is significantly bottle-necked by the explosive inflation of the communication payloads, based on ciphertexts subject to the increasing length of plaintexts. In this way, we then proposed to use stochastic approximation techniques to reduced the communication payloads while accelerating the overall training procedure in a statistical manner. Our experiments using the real-world data showed that SecureGBM can well secure the communication and computation of LightGBM training and inference procedures for the both parties while only losing less than 3% AUC, using the same number of iterations for gradient boosting, on a wide range of benchmark datasets.

Xingjian Li, Haoyi Xiong, Hanchao Wang et al.

Transfer learning through fine-tuning a pre-trained neural network with an extremely large dataset, such as ImageNet, can significantly accelerate training while the accuracy is frequently bottlenecked by the limited dataset size of the new target task. To solve the problem, some regularization methods, constraining the outer layer weights of the target network using the starting point as references (SPAR), have been studied. In this paper, we propose a novel regularized transfer learning framework DELTA, namely DEep Learning Transfer using Feature Map with Attention. Instead of constraining the weights of neural network, DELTA aims to preserve the outer layer outputs of the target network. Specifically, in addition to minimizing the empirical loss, DELTA intends to align the outer layer outputs of two networks, through constraining a subset of feature maps that are precisely selected by attention that has been learned in an supervised learning manner. We evaluate DELTA with the state-of-the-art algorithms, including L2 and L2-SP. The experiment results show that our proposed method outperforms these baselines with higher accuracy for new tasks.

Jiaxuan Zhuo, Zeyu Chen, Jianhuang Lai et al.

Person re-identification (re-id) suffers from a serious occlusion problem when applied to crowded public places. In this paper, we propose to retrieve a full-body person image by using a person image with occlusions. This differs significantly from the conventional person re-id problem where it is assumed that person images are detected without any occlusion. We thus call this new problem the occluded person re-identitification. To address this new problem, we propose a novel Attention Framework of Person Body (AFPB) based on deep learning, consisting of 1) an Occlusion Simulator (OS) which automatically generates artificial occlusions for full-body person images, and 2) multi-task losses that force the neural network not only to discriminate a person's identity but also to determine whether a sample is from the occluded data distribution or the full-body data distribution. Experiments on a new occluded person re-id dataset and three existing benchmarks modified to include full-body person images and occluded person images show the superiority of the proposed method.

Long Jin, Zeyu Chen, Zhuowen Tu

Instance segmentation has attracted recent attention in computer vision and existing methods in this domain mostly have an object detection stage. In this paper, we study the intrinsic challenge of the instance segmentation problem, the presence of a quotient space (swapping the labels of different instances leads to the same result), and propose new methods that are object proposal- and object detection- free. We propose three alternative methods, namely pixel-based affinity mapping, superpixel-based affinity learning, and boundary-based component segmentation, all focusing on performing labeling transformations to cope with the quotient space problem. By adopting fully convolutional neural networks (FCN) like models, our framework attains competitive results on both the PASCAL dataset (object-centric) and the Gland dataset (texture-centric), which the existing methods are not able to do. Our work also has the advantages in its transparency, simplicity, and being all segmentation based.