Weiming Hu

Guanyi Qin, Jie Liang, Bingbing Zhang et al. · baidu

In this paper, we present an overview of the NTIRE 2026 challenge on the 3rd Restore Any Image Model in the Wild, specifically focusing on Track 1: Professional Image Quality Assessment. Conventional Image Quality Assessment (IQA) typically relies on scalar scores. By compressing complex visual characteristics into a single number, these methods fundamentally struggle to distinguish subtle differences among uniformly high-quality images. Furthermore, they fail to articulate why one image is superior, lacking the reasoning capabilities required to provide guidance for vision tasks. To bridge this gap, recent advancements in Multimodal Large Language Models (MLLMs) offer a promising paradigm. Inspired by this potential, our challenge establishes a novel benchmark exploring the ability of MLLMs to mimic human expert cognition in evaluating high-quality image pairs. Participants were tasked with overcoming critical bottlenecks in professional scenarios, centering on two primary objectives: (1) Comparative Quality Selection: reliably identifying the visually superior image within a high-quality pair; and (2) Interpretative Reasoning: generating grounded, expert-level explanations that detail the rationale behind the selection. In total, the challenge attracted nearly 200 registrations and over 2,500 submissions. The top-performing methods significantly advanced the state of the art in professional IQA. The challenge dataset is available at https://github.com/narthchin/RAIM-PIQA, and the official homepage is accessible at https://www.codabench.org/competitions/12789/.

Shaoru Wang, Jin Gao, Zeming Li et al.

Self-supervised learning on large-scale Vision Transformers (ViTs) as pre-training methods has achieved promising downstream performance. Yet, how much these pre-training paradigms promote lightweight ViTs' performance is considerably less studied. In this work, we develop and benchmark several self-supervised pre-training methods on image classification tasks and some downstream dense prediction tasks. We surprisingly find that if proper pre-training is adopted, even vanilla lightweight ViTs show comparable performance to previous SOTA networks with delicate architecture design. It breaks the recently popular conception that vanilla ViTs are not suitable for vision tasks in lightweight regimes. We also point out some defects of such pre-training, e.g., failing to benefit from large-scale pre-training data and showing inferior performance on data-insufficient downstream tasks. Furthermore, we analyze and clearly show the effect of such pre-training by analyzing the properties of the layer representation and attention maps for related models. Finally, based on the above analyses, a distillation strategy during pre-training is developed, which leads to further downstream performance improvement for MAE-based pre-training. Code is available at https://github.com/wangsr126/mae-lite.

Li Yang, Yan Xu, Chunfeng Yuan et al.

Visual grounding is a task to locate the target indicated by a natural language expression. Existing methods extend the generic object detection framework to this problem. They base the visual grounding on the features from pre-generated proposals or anchors, and fuse these features with the text embeddings to locate the target mentioned by the text. However, modeling the visual features from these predefined locations may fail to fully exploit the visual context and attribute information in the text query, which limits their performance. In this paper, we propose a transformer-based framework for accurate visual grounding by establishing text-conditioned discriminative features and performing multi-stage cross-modal reasoning. Specifically, we develop a visual-linguistic verification module to focus the visual features on regions relevant to the textual descriptions while suppressing the unrelated areas. A language-guided feature encoder is also devised to aggregate the visual contexts of the target object to improve the object's distinctiveness. To retrieve the target from the encoded visual features, we further propose a multi-stage cross-modal decoder to iteratively speculate on the correlations between the image and text for accurate target localization. Extensive experiments on five widely used datasets validate the efficacy of our proposed components and demonstrate state-of-the-art performance. Our code is public at https://github.com/yangli18/VLTVG.

Haowei Liu, Xi Zhang, Haiyang Xu et al.

Built on the power of LLMs, numerous multimodal large language models (MLLMs) have recently achieved remarkable performance on various vision-language tasks. However, most existing MLLMs and benchmarks primarily focus on single-image input scenarios, leaving the performance of MLLMs when handling realistic multiple images underexplored. Although a few benchmarks consider multiple images, their evaluation dimensions and samples are very limited. In this paper, we propose a new benchmark MIBench, to comprehensively evaluate fine-grained abilities of MLLMs in multi-image scenarios. Specifically, MIBench categorizes the multi-image abilities into three scenarios: multi-image instruction (MII), multimodal knowledge-seeking (MKS) and multimodal in-context learning (MIC), and constructs 13 tasks with a total of 13K annotated samples. During data construction, for MII and MKS, we extract correct options from manual annotations and create challenging distractors to obtain multiple-choice questions. For MIC, to enable an in-depth evaluation, we set four sub-tasks and transform the original datasets into in-context learning formats. We evaluate several open-source and closed-source MLLMs on the proposed MIBench. The results reveal that although current models excel in single-image tasks, they exhibit significant shortcomings when faced with multi-image inputs, such as limited fine-grained perception, multi-image reasoning and in-context learning abilities. The annotated data of MIBench is available at https://huggingface.co/datasets/StarBottle/MIBench.

Weiming Hu, Qiang Wang, Li Zhang et al.

In this paper we introduce SiamMask, a framework to perform both visual object tracking and video object segmentation, in real-time, with the same simple method. We improve the offline training procedure of popular fully-convolutional Siamese approaches by augmenting their losses with a binary segmentation task. Once the offline training is completed, SiamMask only requires a single bounding box for initialization and can simultaneously carry out visual object tracking and segmentation at high frame-rates. Moreover, we show that it is possible to extend the framework to handle multiple object tracking and segmentation by simply re-using the multi-task model in a cascaded fashion. Experimental results show that our approach has high processing efficiency, at around 55 frames per second. It yields real-time state-of-the-art results on visual-object tracking benchmarks, while at the same time demonstrating competitive performance at a high speed for video object segmentation benchmarks.

Yutong Kou, Jin Gao, Bing Li et al.

Recently, the transformer has enabled the speed-oriented trackers to approach state-of-the-art (SOTA) performance with high-speed thanks to the smaller input size or the lighter feature extraction backbone, though they still substantially lag behind their corresponding performance-oriented versions. In this paper, we demonstrate that it is possible to narrow or even close this gap while achieving high tracking speed based on the smaller input size. To this end, we non-uniformly resize the cropped image to have a smaller input size while the resolution of the area where the target is more likely to appear is higher and vice versa. This enables us to solve the dilemma of attending to a larger visual field while retaining more raw information for the target despite a smaller input size. Our formulation for the non-uniform resizing can be efficiently solved through quadratic programming (QP) and naturally integrated into most of the crop-based local trackers. Comprehensive experiments on five challenging datasets based on two kinds of transformer trackers, \ie, OSTrack and TransT, demonstrate consistent improvements over them. In particular, applying our method to the speed-oriented version of OSTrack even outperforms its performance-oriented counterpart by 0.6% AUC on TNL2K, while running 50% faster and saving over 55% MACs. Codes and models are available at https://github.com/Kou-99/ZoomTrack.

Wanli Liu, Chen Li, Ning Xu et al.

Cervical cancer is the seventh most common cancer among all the cancers worldwide and the fourth most common cancer among women. Cervical cytopathology image classification is an important method to diagnose cervical cancer. Manual screening of cytopathology images is time-consuming and error-prone. The emergence of the automatic computer-aided diagnosis system solves this problem. This paper proposes a framework called CVM-Cervix based on deep learning to perform cervical cell classification tasks. It can analyze pap slides quickly and accurately. CVM-Cervix first proposes a Convolutional Neural Network module and a Visual Transformer module for local and global feature extraction respectively, then a Multilayer Perceptron module is designed to fuse the local and global features for the final classification. Experimental results show the effectiveness and potential of the proposed CVM-Cervix in the field of cervical Pap smear image classification. In addition, according to the practical needs of clinical work, we perform a lightweight post-processing to compress the model.

Yunfei Zhang, Chao Liang, Jin Gao et al.

Joint Detection and Embedding (JDE) trackers have demonstrated excellent performance in Multi-Object Tracking (MOT) tasks by incorporating the extraction of appearance features as auxiliary tasks through embedding Re-Identification task (ReID) into the detector, achieving a balance between inference speed and tracking performance. However, solving the competition between the detector and the feature extractor has always been a challenge. Meanwhile, the issue of directly embedding the ReID task into MOT has remained unresolved. The lack of high discriminability in appearance features results in their limited utility. In this paper, a new learning approach using cross-correlation to capture temporal information of objects is proposed. The feature extraction network is no longer trained solely on appearance features from each frame but learns richer motion features by utilizing feature heatmaps from consecutive frames, which addresses the challenge of inter-class feature similarity. Furthermore, our learning approach is applied to a more lightweight feature extraction network, and treat the feature matching scores as strong cues rather than auxiliary cues, with an appropriate weight calculation to reflect the compatibility between our obtained features and the MOT task. Our tracker, named TCBTrack, achieves state-of-the-art performance on multiple public benchmarks, i.e., MOT17, MOT20, and DanceTrack datasets. Specifically, on the DanceTrack test set, we achieve 56.8 HOTA, 58.1 IDF1 and 92.5 MOTA, making it the best online tracker capable of achieving real-time performance. Comparative evaluations with other trackers prove that our tracker achieves the best balance between speed, robustness and accuracy. Code is available at https://github.com/yfzhang1214/TCBTrack.

Liyu Shi, Xiaoyan Li, Weiming Hu et al.

Background and Purpose: Colorectal cancer is a common fatal malignancy, the fourth most common cancer in men, and the third most common cancer in women worldwide. Timely detection of cancer in its early stages is essential for treating the disease. Currently, there is a lack of datasets for histopathological image segmentation of rectal cancer, which often hampers the assessment accuracy when computer technology is used to aid in diagnosis. Methods: This present study provided a new publicly available Enteroscope Biopsy Histopathological Hematoxylin and Eosin Image Dataset for Image Segmentation Tasks (EBHI-Seg). To demonstrate the validity and extensiveness of EBHI-Seg, the experimental results for EBHI-Seg are evaluated using classical machine learning methods and deep learning methods. Results: The experimental results showed that deep learning methods had a better image segmentation performance when utilizing EBHI-Seg. The maximum accuracy of the Dice evaluation metric for the classical machine learning method is 0.948, while the Dice evaluation metric for the deep learning method is 0.965. Conclusion: This publicly available dataset contained 5,170 images of six types of tumor differentiation stages and the corresponding ground truth images. The dataset can provide researchers with new segmentation algorithms for medical diagnosis of colorectal cancer, which can be used in the clinical setting to help doctors and patients.

Haoyuan Chen, Chen Li, Xiaoyan Li et al.

In recent years, colorectal cancer has become one of the most significant diseases that endanger human health. Deep learning methods are increasingly important for the classification of colorectal histopathology images. However, existing approaches focus more on end-to-end automatic classification using computers rather than human-computer interaction. In this paper, we propose an IL-MCAM framework. It is based on attention mechanisms and interactive learning. The proposed IL-MCAM framework includes two stages: automatic learning (AL) and interactivity learning (IL). In the AL stage, a multi-channel attention mechanism model containing three different attention mechanism channels and convolutional neural networks is used to extract multi-channel features for classification. In the IL stage, the proposed IL-MCAM framework continuously adds misclassified images to the training set in an interactive approach, which improves the classification ability of the MCAM model. We carried out a comparison experiment on our dataset and an extended experiment on the HE-NCT-CRC-100K dataset to verify the performance of the proposed IL-MCAM framework, achieving classification accuracies of 98.98% and 99.77%, respectively. In addition, we conducted an ablation experiment and an interchangeability experiment to verify the ability and interchangeability of the three channels. The experimental results show that the proposed IL-MCAM framework has excellent performance in the colorectal histopathological image classification tasks.

Zongyang Ma, Guan Luo, Jin Gao et al.

Open-vocabulary object detection aims to detect novel object categories beyond the training set. The advanced open-vocabulary two-stage detectors employ instance-level visual-to-visual knowledge distillation to align the visual space of the detector with the semantic space of the Pre-trained Visual-Language Model (PVLM). However, in the more efficient one-stage detector, the absence of class-agnostic object proposals hinders the knowledge distillation on unseen objects, leading to severe performance degradation. In this paper, we propose a hierarchical visual-language knowledge distillation method, i.e., HierKD, for open-vocabulary one-stage detection. Specifically, a global-level knowledge distillation is explored to transfer the knowledge of unseen categories from the PVLM to the detector. Moreover, we combine the proposed global-level knowledge distillation and the common instance-level knowledge distillation to learn the knowledge of seen and unseen categories simultaneously. Extensive experiments on MS-COCO show that our method significantly surpasses the previous best one-stage detector with 11.9\% and 6.7\% $AP_{50}$ gains under the zero-shot detection and generalized zero-shot detection settings, and reduces the $AP_{50}$ performance gap from 14\% to 7.3\% compared to the best two-stage detector.

Yufan Liu, Jiajiong Cao, Bing Li et al.

Transformer attracts much attention because of its ability to learn global relations and superior performance. In order to achieve higher performance, it is natural to distill complementary knowledge from Transformer to convolutional neural network (CNN). However, most existing knowledge distillation methods only consider homologous-architecture distillation, such as distilling knowledge from CNN to CNN. They may not be suitable when applying to cross-architecture scenarios, such as from Transformer to CNN. To deal with this problem, a novel cross-architecture knowledge distillation method is proposed. Specifically, instead of directly mimicking output/intermediate features of the teacher, partially cross attention projector and group-wise linear projector are introduced to align the student features with the teacher's in two projected feature spaces. And a multi-view robust training scheme is further presented to improve the robustness and stability of the framework. Extensive experiments show that the proposed method outperforms 14 state-of-the-arts on both small-scale and large-scale datasets.

Jiale Xu, Rui Zhang, Yi Xiong et al.

Large Language Models are increasingly being deployed in datacenters. Serving these models requires careful memory management, as their memory usage includes static weights, dynamic activations, and key-value caches. While static weights are constant and predictable, dynamic components such as activations and KV caches change frequently during runtime, presenting significant challenges for efficient memory management. Modern LLM serving systems typically handle runtime memory and KV caches at distinct abstraction levels: runtime memory management relies on static tensor abstractions, whereas KV caches utilize a page table-based virtualization layer built on top of the tensor abstraction. This virtualization dynamically manages KV caches to mitigate memory fragmentation. However, this dual-level approach fundamentally isolates runtime memory and KV cache management, resulting in suboptimal memory utilization under dynamic workloads, which can lead to a nearly 20% drop in throughput. To address these limitations, we propose eLLM, an elastic memory management framework inspired by the classical memory ballooning mechanism in operating systems. The core components of eLLM include: (1) Virtual Tensor Abstraction, which decouples the virtual address space of tensors from the physical GPU memory, creating a unified and flexible memory pool; (2) an Elastic Memory Mechanism that dynamically adjusts memory allocation through runtime memory inflation and deflation, leveraging CPU memory as an extensible buffer; and (3) a Lightweight Scheduling Strategy employing SLO-aware policies to optimize memory utilization and effectively balance performance trade-offs under stringent SLO constraints. Comprehensive evaluations demonstrate that eLLM significantly outperforms state-of-the-art systems, 2.32x higher decoding throughput, and supporting 3x larger batch sizes for 128K-token inputs.

Yanqin Jiang, Li Zhang, Zhenwei Miao et al.

3D object detection in autonomous driving aims to reason "what" and "where" the objects of interest present in a 3D world. Following the conventional wisdom of previous 2D object detection, existing methods often adopt the canonical Cartesian coordinate system with perpendicular axis. However, we conjugate that this does not fit the nature of the ego car's perspective, as each onboard camera perceives the world in shape of wedge intrinsic to the imaging geometry with radical (non-perpendicular) axis. Hence, in this paper we advocate the exploitation of the Polar coordinate system and propose a new Polar Transformer (PolarFormer) for more accurate 3D object detection in the bird's-eye-view (BEV) taking as input only multi-camera 2D images. Specifically, we design a cross attention based Polar detection head without restriction to the shape of input structure to deal with irregular Polar grids. For tackling the unconstrained object scale variations along Polar's distance dimension, we further introduce a multi-scalePolar representation learning strategy. As a result, our model can make best use of the Polar representation rasterized via attending to the corresponding image observation in a sequence-to-sequence fashion subject to the geometric constraints. Thorough experiments on the nuScenes dataset demonstrate that our PolarFormer outperforms significantly state-of-the-art 3D object detection alternatives.

Weiming Hu, Haoyuan Chen, Wanli Liu et al.

Gastric cancer is the fifth most common cancer in the world. At the same time, it is also the fourth most deadly cancer. Early detection of cancer exists as a guide for the treatment of gastric cancer. Nowadays, computer technology has advanced rapidly to assist physicians in the diagnosis of pathological pictures of gastric cancer. Ensemble learning is a way to improve the accuracy of algorithms, and finding multiple learning models with complementarity types is the basis of ensemble learning. The complementarity of sub-size pathology image classifiers when machine performance is insufficient is explored in this experimental platform. We choose seven classical machine learning classifiers and four deep learning classifiers for classification experiments on the GasHisSDB database. Among them, classical machine learning algorithms extract five different image virtual features to match multiple classifier algorithms. For deep learning, we choose three convolutional neural network classifiers. In addition, we also choose a novel Transformer-based classifier. The experimental platform, in which a large number of classical machine learning and deep learning methods are performed, demonstrates that there are differences in the performance of different classifiers on GasHisSDB. Classical machine learning models exist for classifiers that classify Abnormal categories very well, while classifiers that excel in classifying Normal categories also exist. Deep learning models also exist with multiple models that can be complementarity. Suitable classifiers are selected for ensemble learning, when machine performance is insufficient. This experimental platform demonstrates that multiple classifiers are indeed complementarity and can improve the efficiency of ensemble learning. This can better assist doctors in diagnosis, improve the detection of gastric cancer, and increase the cure rate.

Zewen Chen, Sunhan Xu, Yun Zeng et al.

With the rising demand for high-resolution (HR) images, No-Reference Image Quality Assessment (NR-IQA) gains more attention, as it can ecaluate image quality in real-time on mobile devices and enhance user experience. However, existing NR-IQA methods often resize or crop the HR images into small resolution, which leads to a loss of important details. And most of them are of high computational complexity, which hinders their application on mobile devices due to limited computational resources. To address these challenges, we propose MobileIQA, a novel approach that utilizes lightweight backbones to efficiently assess image quality while preserving image details through high-resolution input. MobileIQA employs the proposed multi-view attention learning (MAL) module to capture diverse opinions, simulating subjective opinions provided by different annotators during the dataset annotation process. The model uses a teacher model to guide the learning of a student model through knowledge distillation. This method significantly reduces computational complexity while maintaining high performance. Experiments demonstrate that MobileIQA outperforms novel IQA methods on evaluation metrics and computational efficiency. The code is available at https://github.com/chencn2020/MobileIQA.

Shaoru Wang, Jin Gao, Bing Li et al.

Deep learning methods require massive of annotated data for optimizing parameters. For example, datasets attached with accurate bounding box annotations are essential for modern object detection tasks. However, labeling with such pixel-wise accuracy is laborious and time-consuming, and elaborate labeling procedures are indispensable for reducing man-made noise, involving annotation review and acceptance testing. In this paper, we focus on the impact of noisy location annotations on the performance of object detection approaches and aim to, on the user side, reduce the adverse effect of the noise. First, noticeable performance degradation is experimentally observed for both one-stage and two-stage detectors when noise is introduced to the bounding box annotations. For instance, our synthesized noise results in performance decrease from 38.9% AP to 33.6% AP for FCOS detector on COCO test split, and 37.8%AP to 33.7%AP for Faster R-CNN. Second, a self-correction technique based on a Bayesian filter for prediction ensemble is proposed to better exploit the noisy location annotations following a Teacher-Student learning paradigm. Experiments for both synthesized and real-world scenarios consistently demonstrate the effectiveness of our approach, e.g., our method increases the degraded performance of the FCOS detector from 33.6% AP to 35.6% AP on COCO.

Jiale Xu, Rui Zhang, Cong Guo et al.

Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests. This surge in demand poses significant challenges in optimizing throughput and latency while keeping costs manageable. The Key-Value (KV) cache, a standard method for retaining previous computations, makes LLM inference highly bounded by memory. While batching strategies can enhance performance, they frequently lead to significant memory fragmentation. Even though cutting-edge systems like vLLM mitigate KV cache fragmentation using paged Attention mechanisms, they still suffer from inefficient memory and computational operations due to the tightly coupled page management and computation kernels. This study introduces the vTensor, an innovative tensor structure for LLM inference based on GPU virtual memory management (VMM). vTensor addresses existing limitations by decoupling computation from memory defragmentation and offering dynamic extensibility. Our framework employs a CPU-GPU heterogeneous approach, ensuring efficient, fragmentation-free memory management while accommodating various computation kernels across different LLM architectures. Experimental results indicate that vTensor achieves an average speedup of 1.86x across different models, with up to 2.42x in multi-turn chat scenarios. Additionally, vTensor provides average speedups of 2.12x and 3.15x in kernel evaluation, reaching up to 3.92x and 3.27x compared to SGLang Triton prefix-prefilling kernels and vLLM paged Attention kernel, respectively. Furthermore, it frees approximately 71.25% (57GB) of memory on the NVIDIA A100 GPU compared to vLLM, enabling more memory-intensive workloads.

Yiwei Zhang, Xuesong Chen, Jin Gao et al.

Vision-Language Models(VLMs) excel at autoregressive text generation, yet end-to-end autonomous driving requires multi-task learning with structured outputs and heterogeneous decoding behaviors, such as autoregressive language generation, parallel object detection and trajectory regression. To accommodate these differences, existing systems typically introduce separate or cascaded decoders, resulting in architectural fragmentation and limited backbone reuse. In this work, we present a unified autonomous driving framework built upon a pretrained VLM, where heterogeneous decoding behaviors are reconciled within a single transformer decoder. We demonstrate that pretrained VLM attention exhibits strong transferability beyond pure language modeling. By organizing visual and structured query tokens within a single causal decoder, structured queries can naturally condition on visual context through the original attention mechanism. Textual and structured outputs share a common attention backbone, enabling stable joint optimization across heterogeneous tasks. Trajectory planning is realized within the same causal LLM decoder by introducing structured trajectory queries. This unified formulation enables planning to share the pretrained attention backbone with images and perception tokens. Extensive experiments on end-to-end autonomous driving benchmarks demonstrate state-of-the-art performance, including 0.28 L2 and 0.18 collision rate on nuScenes open-loop evaluation and competitive results (86.8 PDMS) on NAVSIM closed-loop evaluation. The full model preserves multi-modal generation capability, while an efficient inference mode achieves approximately 40% lower latency. Code and models are available at https://github.com/Z1zyw/OneDrive

Yanqin Jiang, Chaohui Yu, Chenjie Cao et al.

Recent advances in 4D generation mainly focus on generating 4D content by distilling pre-trained text or single-view image-conditioned models. It is inconvenient for them to take advantage of various off-the-shelf 3D assets with multi-view attributes, and their results suffer from spatiotemporal inconsistency owing to the inherent ambiguity in the supervision signals. In this work, we present Animate3D, a novel framework for animating any static 3D model. The core idea is two-fold: 1) We propose a novel multi-view video diffusion model (MV-VDM) conditioned on multi-view renderings of the static 3D object, which is trained on our presented large-scale multi-view video dataset (MV-Video). 2) Based on MV-VDM, we introduce a framework combining reconstruction and 4D Score Distillation Sampling (4D-SDS) to leverage the multi-view video diffusion priors for animating 3D objects. Specifically, for MV-VDM, we design a new spatiotemporal attention module to enhance spatial and temporal consistency by integrating 3D and video diffusion models. Additionally, we leverage the static 3D model's multi-view renderings as conditions to preserve its identity. For animating 3D models, an effective two-stage pipeline is proposed: we first reconstruct motions directly from generated multi-view videos, followed by the introduced 4D-SDS to refine both appearance and motion. Benefiting from accurate motion learning, we could achieve straightforward mesh animation. Qualitative and quantitative experiments demonstrate that Animate3D significantly outperforms previous approaches. Data, code, and models will be open-released.

Yufan Liu, Daoyuan Ren, Zhipeng Zhang et al.

Feature compression is increasingly important for improving the efficiency of downstream tasks, especially in applications involving large-scale or multi-modal data. While existing methods typically rely on dedicated models for achieving specific compression ratios, they are often limited in flexibility and generalization. In particular, retraining is necessary when adapting to a new compression ratio. To address this limitation, we propose a novel and flexible Arbitrary Ratio Feature Compression (ARFC) framework, which supports any compression ratio with a single model, eliminating the need for multiple specialized models. At its core, the Arbitrary Ratio Compressor (ARC) is an auto-regressive model that performs compression via next-token prediction. This allows the compression ratio to be controlled at inference simply by adjusting the number of generated tokens. To enhance the quality of the compressed features, two key modules are introduced. The Mixture of Solutions (MoS) module refines the compressed tokens by utilizing multiple compression results (solutions), reducing uncertainty and improving robustness. The Entity Relation Graph Constraint (ERGC) is integrated into the training process to preserve semantic and structural relationships during compression. Extensive experiments on cross-modal retrieval, image classification, and image retrieval tasks across multiple datasets demonstrate that our method consistently outperforms existing approaches at various compression ratios. Notably, in some cases, it even surpasses the performance of the original, uncompressed features. These results validate the effectiveness and versatility of ARFC for practical, resource-constrained scenarios.

Zongyang Ma, Ziqi Zhang, Yuxin Chen et al.

Understanding the content of events occurring in the video and their inherent temporal logic is crucial for video-text retrieval. However, web-crawled pre-training datasets often lack sufficient event information, and the widely adopted video-level cross-modal contrastive learning also struggles to capture detailed and complex video-text event alignment. To address these challenges, we make improvements from both data and model perspectives. In terms of pre-training data, we focus on supplementing the missing specific event content and event temporal transitions with the proposed event augmentation strategies. Based on the event-augmented data, we construct a novel Event-Aware Video-Text Retrieval model, ie, EA-VTR, which achieves powerful video-text retrieval ability through superior video event awareness. EA-VTR can efficiently encode frame-level and video-level visual representations simultaneously, enabling detailed event content and complex event temporal cross-modal alignment, ultimately enhancing the comprehensive understanding of video events. Our method not only significantly outperforms existing approaches on multiple datasets for Text-to-Video Retrieval and Video Action Recognition tasks, but also demonstrates superior event content perceive ability on Multi-event Video-Text Retrieval and Video Moment Retrieval tasks, as well as outstanding event temporal logic understanding ability on Test of Time task.

Yuxin Chen, Zongyang Ma, Ziqi Zhang et al.

Dominant dual-encoder models enable efficient image-text retrieval but suffer from limited accuracy while the cross-encoder models offer higher accuracy at the expense of efficiency. Distilling cross-modality matching knowledge from cross-encoder to dual-encoder provides a natural approach to harness their strengths. Thus we investigate the following valuable question: how to make cross-encoder a good teacher for dual-encoder? Our findings are threefold:(1) Cross-modal similarity score distribution of cross-encoder is more concentrated while the result of dual-encoder is nearly normal making vanilla logit distillation less effective. However ranking distillation remains practical as it is not affected by the score distribution.(2) Only the relative order between hard negatives conveys valid knowledge while the order information between easy negatives has little significance.(3) Maintaining the coordination between distillation loss and dual-encoder training loss is beneficial for knowledge transfer. Based on these findings we propose a novel Contrastive Partial Ranking Distillation (CPRD) method which implements the objective of mimicking relative order between hard negative samples with contrastive learning. This approach coordinates with the training of the dual-encoder effectively transferring valid knowledge from the cross-encoder to the dual-encoder. Extensive experiments on image-text retrieval and ranking tasks show that our method surpasses other distillation methods and significantly improves the accuracy of dual-encoder.

Yifan Lu, Ziqi Zhang, Yuxin Chen et al.

The task of Dense Video Captioning (DVC) aims to generate captions with timestamps for multiple events in one video. Semantic information plays an important role for both localization and description of DVC. We present a semantic-assisted dense video captioning model based on the encoding-decoding framework. In the encoding stage, we design a concept detector to extract semantic information, which is then fused with multi-modal visual features to sufficiently represent the input video. In the decoding stage, we design a classification head, paralleled with the localization and captioning heads, to provide semantic supervision. Our method achieves significant improvements on the YouMakeup dataset under DVC evaluation metrics and achieves high performance in the Makeup Dense Video Captioning (MDVC) task of PIC 4th Challenge.

Cheng Feng, Congxuan Zhang, Zhen Chen et al.

Depth sensing is of paramount importance for unmanned aerial and autonomous vehicles. Nonetheless, contemporary monocular depth estimation methods employing complex deep neural networks within Convolutional Neural Networks are inadequately expedient for real-time inference on embedded platforms. This paper endeavors to surmount this challenge by proposing two efficient and lightweight architectures, RT-MonoDepth and RT-MonoDepth-S, thereby mitigating computational complexity and latency. Our methodologies not only attain accuracy comparable to prior depth estimation methods but also yield faster inference speeds. Specifically, RT-MonoDepth and RT-MonoDepth-S achieve frame rates of 18.4&30.5 FPS on NVIDIA Jetson Nano and 253.0&364.1 FPS on Jetson AGX Orin, utilizing a single RGB image of resolution 640x192. The experimental results underscore the superior accuracy and faster inference speed of our methods in comparison to existing fast monocular depth estimation methodologies on the KITTI dataset.

Shaoru Wang, Zeming Li, Jin Gao et al.

Self-supervised learning (SSL) has achieved promising downstream performance. However, when facing various resource budgets in real-world applications, it costs a huge computation burden to pretrain multiple networks of various sizes one by one. In this paper, we propose Discriminative-SSL-based Slimmable Pretrained Networks (DSPNet), which can be trained at once and then slimmed to multiple sub-networks of various sizes, each of which faithfully learns good representation and can serve as good initialization for downstream tasks with various resource budgets. Specifically, we extend the idea of slimmable networks to a discriminative SSL paradigm, by integrating SSL and knowledge distillation gracefully. We show comparable or improved performance of DSPNet on ImageNet to the networks individually pretrained one by one under the linear evaluation and semi-supervised evaluation protocols, while reducing large training cost. The pretrained models also generalize well on downstream detection and segmentation tasks. Code will be made public.

Junbin Su, Ziteng Xue, Shihui Zhang et al.

Parameter-efficient fine-tuning (PEFT) in multimodal tracking reveals a concerning trend where recent performance gains are often achieved at the cost of inflated parameter budgets, which fundamentally erodes PEFT's efficiency promise. In this work, we introduce SEATrack, a Simple, Efficient, and Adaptive two-stream multimodal tracker that tackles this performance-efficiency dilemma from two complementary perspectives. We first prioritize cross-modal alignment of matching responses, an underexplored yet pivotal factor that we argue is essential for breaking the trade-off. Specifically, we observe that modality-specific biases in existing two-stream methods generate conflicting matching attention maps, thereby hindering effective joint representation learning. To mitigate this, we propose AMG-LoRA, which seamlessly integrates Low-Rank Adaptation (LoRA) for domain adaptation with Adaptive Mutual Guidance (AMG) to dynamically refine and align attention maps across modalities. We then depart from conventional local fusion approaches by introducing a Hierarchical Mixture of Experts (HMoE) that enables efficient global relation modeling, effectively balancing expressiveness and computational efficiency in cross-modal fusion. Equipped with these innovations, SEATrack advances notable progress over state-of-the-art methods in balancing performance with efficiency across RGB-T, RGB-D, and RGB-E tracking tasks. \href{https://github.com/AutoLab-SAI-SJTU/SEATrack}{\textcolor{cyan}{Code is available}}.

Yuxin Yang, Yinan Zhou, Yuxin Chen et al.

Composed Image Retrieval (CIR) has demonstrated significant potential by enabling flexible multimodal queries that combine a reference image and modification text. However, CIR inherently prioritizes semantic matching, struggling to reliably retrieve a user-specified instance across contexts. In practice, emphasizing concrete instance fidelity over broad semantics is often more consequential. In this work, we propose Object-Anchored Composed Image Retrieval (OACIR), a novel fine-grained retrieval task that mandates strict instance-level consistency. To advance research on this task, we construct OACIRR (OACIR on Real-world images), the first large-scale, multi-domain benchmark comprising over 160K quadruples and four challenging candidate galleries enriched with hard-negative instance distractors. Each quadruple augments the compositional query with a bounding box that visually anchors the object in the reference image, providing a precise and flexible way to ensure instance preservation. To address the OACIR task, we propose AdaFocal, a framework featuring a Context-Aware Attention Modulator that adaptively intensifies attention within the specified instance region, dynamically balancing focus between the anchored instance and the broader compositional context. Extensive experiments demonstrate that AdaFocal substantially outperforms existing compositional retrieval models, particularly in maintaining instance-level fidelity, thereby establishing a robust baseline for this challenging task while opening new directions for more flexible, instance-aware retrieval systems.

Zhiheng Li, Zongyang Ma, Jiaxian Chen et al.

The past year has seen over 20 open-source document parsing models, yet thefield still benchmarks almost exclusively on OmniDocBench, a 1,355-pagemanually annotated dataset whose top scores have saturated above 90%. Athree-stage audit pipeline we run on OmniDocBench screens its 21,353evaluator-scored blocks and confirms 2,580 errors (12.08%); combined with overa year of public availability, both annotation quality and contamination riskcall its rankings into question. To address these issues, we presentPureDocBench, a programmatically generated, source-traceable benchmark thatrenders document images from HTML/CSS and produces verifiable annotations fromthe same source, covering 10 domains, 66 subcategories, and 1,475 pages, eachin three versions: clean, digitally degraded, and real-degraded (4,425 imagestotal). Evaluating 40 models spanning pipeline specialists, end-to-endspecialists, and general-purpose VLMs, we find: (i) document parsing is farfrom solved: the best model scores only ~74 out of 100, with a 44.6-point gapbetween the strongest and weakest models; (ii) specialist parsers with <=4Bparameters rival or surpass general VLMs that are 5-100x larger, yet formularecognition remains a shared bottleneck where no model exceeds 67% whenaveraging the formula metric across all three tracks; (iii) general VLMs loseonly 0.99/8.52 Overall points under digital/real degradation versus 4.90/14.21for pipeline specialists, producing ranking reversals that make clean-onlyevaluation misleading for deployment. All data, code, and artifacts arepublicly released.

Jinming Lou, Wenyang Luo, Yufan Liu et al.

Diffusion transformers have gained substantial interest in diffusion generative modeling due to their outstanding performance. However, their computational demands, particularly the quadratic complexity of attention mechanisms and multi-step inference processes, present substantial bottlenecks that limit their practical applications. To address these challenges, we propose TokenCache, a novel acceleration method that leverages the token-based multi-block architecture of transformers to reduce redundant computations. TokenCache tackles three critical questions: (1) Which tokens should be pruned and reused by the caching mechanism to eliminate redundancy? (2) Which blocks should be targeted for efficient caching? (3) At which time steps should caching be applied to balance speed and quality? In response to these challenges, TokenCache introduces a Cache Predictor that hierarchically addresses these issues by (1) Token pruning: assigning importance scores to each token to determine which tokens to prune and reuse; (2) Block selection: allocating pruning ratio to each block to adaptively select blocks for caching; (3) Temporal Scheduling: deciding at which time steps to apply caching strategies. Experimental results across various models demonstrate that TokenCache achieves an effective trade-off between generation quality and inference speed for diffusion transformers.

Yue Guan, Changming Yu, Shihan Fang et al.

Speculative decoding improves LLM inference by generating and verifying multiple tokens in parallel, but existing systems suffer from suboptimal performance due to a mismatch between dynamic speculation and static runtime assumptions. We present Yggdrasil, a co-designed system that enables latency-optimal speculative decoding through context-aware tree drafting and compiler-friendly execution. Yggdrasil introduces an equal-growth tree structure for static graph compatibility, a latency-aware optimization objective for draft selection, and stage-based scheduling to reduce overhead. Yggdrasil supports unmodified LLMs and achieves up to $3.98\times$ speedup over state-of-the-art baselines across multiple hardware setups.

Xun Guo, Mingwu Zheng, Liang Hou et al.

Text-guided image-to-video (I2V) generation aims to generate a coherent video that preserves the identity of the input image and semantically aligns with the input prompt. Existing methods typically augment pretrained text-to-video (T2V) models by either concatenating the image with noised video frames channel-wise before being fed into the model or injecting the image embedding produced by pretrained image encoders in cross-attention modules. However, the former approach often necessitates altering the fundamental weights of pretrained T2V models, thus restricting the model's compatibility within the open-source communities and disrupting the model's prior knowledge. Meanwhile, the latter typically fails to preserve the identity of the input image. We present I2V-Adapter to overcome such limitations. I2V-Adapter adeptly propagates the unnoised input image to subsequent noised frames through a cross-frame attention mechanism, maintaining the identity of the input image without any changes to the pretrained T2V model. Notably, I2V-Adapter only introduces a few trainable parameters, significantly alleviating the training cost and also ensures compatibility with existing community-driven personalized models and control tools. Moreover, we propose a novel Frame Similarity Prior to balance the motion amplitude and the stability of generated videos through two adjustable control coefficients. Our experimental results demonstrate that I2V-Adapter is capable of producing high-quality videos. This performance, coupled with its agility and adaptability, represents a substantial advancement in the field of I2V, particularly for personalized and controllable applications.

Tao Zhang, Ziqi Zhang, Zongyang Ma et al.

The ability to perform multi-modal multi-hop reasoning by iteratively integrating information across various modalities and external knowledge is critical for addressing complex real-world challenges. However, existing Multi-modal Large Language Models (MLLMs) are predominantly limited to single-step reasoning, as existing benchmarks lack the complexity needed to evaluate and drive multi-hop abilities. To bridge this gap, we introduce MMhops, a novel, large-scale benchmark designed to systematically evaluate and foster multi-modal multi-hop reasoning. MMhops dataset comprises two challenging task formats, Bridging and Comparison, which necessitate that models dynamically construct complex reasoning chains by integrating external knowledge. To tackle the challenges posed by MMhops, we propose MMhops-R1, a novel multi-modal Retrieval-Augmented Generation (mRAG) framework for dynamic reasoning. Our framework utilizes reinforcement learning to optimize the model for autonomously planning reasoning paths, formulating targeted queries, and synthesizing multi-level information. Comprehensive experiments demonstrate that MMhops-R1 significantly outperforms strong baselines on MMhops, highlighting that dynamic planning and multi-modal knowledge integration are crucial for complex reasoning. Moreover, MMhops-R1 demonstrates strong generalization to tasks requiring fixed-hop reasoning, underscoring the robustness of our dynamic planning approach. In conclusion, our work contributes a challenging new benchmark and a powerful baseline model, and we will release the associated code, data, and weights to catalyze future research in this critical area.

Manyi Li, Yufan Liu, Lai Jiang et al.

Although unlearning-based defenses claim to purge Not-Safe-For-Work (NSFW) concepts from diffusion models (DMs), we reveals that this "forgetting" is largely an illusion. Unlearning partially disrupts the mapping between linguistic symbols and the underlying knowledge, which remains intact as dormant memories. We find that the distributional discrepancy in the denoising process serves as a measurable indicator of how much of the mapping is retained, also reflecting the strength of unlearning. Inspired by this, we propose IVO (Initial Latent Variable Optimization), a concise and powerful attack framework that reactivates these dormant memories by reconstructing the broken mappings. Through Image Inversion}, Adversarial Optimization and Reused Attack, IVO optimizes initial latent variables to realign the noise distribution of unlearned models with their original unsafe states. Extensive experiments across 8 widely used unlearning techniques demonstrate that IVO achieves superior attack success rates and strong semantic consistency, exposing fundamental flaws in current defenses. The code is available at anonymous.4open.science/r/IVO/. Warning: This paper has unsafe images that may offend some readers.

Haowei Liu, Xi Zhang, Haiyang Xu et al.

In the field of MLLM-based GUI agents, compared to smartphones, the PC scenario not only features a more complex interactive environment, but also involves more intricate intra- and inter-app workflows. To address these issues, we propose a hierarchical agent framework named PC-Agent. Specifically, from the perception perspective, we devise an Active Perception Module (APM) to overcome the inadequate abilities of current MLLMs in perceiving screenshot content. From the decision-making perspective, to handle complex user instructions and interdependent subtasks more effectively, we propose a hierarchical multi-agent collaboration architecture that decomposes decision-making processes into Instruction-Subtask-Action levels. Within this architecture, three agents (i.e., Manager, Progress and Decision) are set up for instruction decomposition, progress tracking and step-by-step decision-making respectively. Additionally, a Reflection agent is adopted to enable timely bottom-up error feedback and adjustment. We also introduce a new benchmark PC-Eval with 25 real-world complex instructions. Empirical results on PC-Eval show that our PC-Agent achieves a 32% absolute improvement of task success rate over previous state-of-the-art methods. The code is available at https://github.com/X-PLUG/MobileAgent/tree/main/PC-Agent.

Zewen Chen, Juan Wang, Wen Wang et al.

Existing Image Quality Assessment (IQA) methods achieve remarkable success in analyzing quality for overall image, but few works explore quality analysis for Regions of Interest (ROIs). The quality analysis of ROIs can provide fine-grained guidance for image quality improvement and is crucial for scenarios focusing on region-level quality. This paper proposes a novel network, SEAGULL, which can SEe and Assess ROIs quality with GUidance from a Large vision-Language model. SEAGULL incorporates a vision-language model (VLM), masks generated by Segment Anything Model (SAM) to specify ROIs, and a meticulously designed Mask-based Feature Extractor (MFE) to extract global and local tokens for specified ROIs, enabling accurate fine-grained IQA for ROIs. Moreover, this paper constructs two ROI-based IQA datasets, SEAGULL-100w and SEAGULL-3k, for training and evaluating ROI-based IQA. SEAGULL-100w comprises about 100w synthetic distortion images with 33 million ROIs for pre-training to improve the model's ability of regional quality perception, and SEAGULL-3k contains about 3k authentic distortion ROIs to enhance the model's ability to perceive real world distortions. After pre-training on SEAGULL-100w and fine-tuning on SEAGULL-3k, SEAGULL shows remarkable performance on fine-grained ROI quality assessment. Code and datasets are publicly available at the https://github.com/chencn2020/Seagull.

Zhiheng Li, Zongyang Ma, Yuntong Pan et al.

Multimodal Large Language Models (MLLMs) are increasingly being deployed as automated content moderators. Within this landscape, we uncover a critical threat: Adversarial Smuggling Attacks. Unlike adversarial perturbations (for misclassification) and adversarial jailbreaks (for harmful output generation), adversarial smuggling exploits the Human-AI capability gap. It encodes harmful content into human-readable visual formats that remain AI-unreadable, thereby evading automated detection and enabling the dissemination of harmful content. We classify smuggling attacks into two pathways: (1) Perceptual Blindness, disrupting text recognition; and (2) Reasoning Blockade, inhibiting semantic understanding despite successful text recognition. To evaluate this threat, we constructed SmuggleBench, the first comprehensive benchmark comprising 1,700 adversarial smuggling attack instances. Evaluations on SmuggleBench reveal that both proprietary (e.g., GPT-5) and open-source (e.g., Qwen3-VL) state-of-the-art models are vulnerable to this threat, producing Attack Success Rates (ASR) exceeding 90%. By analyzing the vulnerability through the lenses of perception and reasoning, we identify three root causes: the limited capabilities of vision encoders, the robustness gap in OCR, and the scarcity of domain-specific adversarial examples. We conduct a preliminary exploration of mitigation strategies, investigating the potential of test-time scaling (via CoT) and adversarial training (via SFT) to mitigate this threat. Our code is publicly available at https://github.com/zhihengli-casia/smugglebench.

Fudong Ge, Yiwei Zhang, Shuhan Shen et al.

In this paper, we propose a new image-based visual place recognition (VPR) framework by exploiting the structural cues in bird's-eye view (BEV) from a single monocular camera. The motivation arises from two key observations about place recognition methods based on both appearance and structure: 1) For the methods relying on LiDAR sensors, the integration of LiDAR in robotic systems has led to increased expenses, while the alignment of data between different sensors is also a major challenge. 2) Other image-/camera-based methods, involving integrating RGB images and their derived variants (eg, pseudo depth images, pseudo 3D point clouds), exhibit several limitations, such as the failure to effectively exploit the explicit spatial relationships between different objects. To tackle the above issues, we design a new BEV-enhanced VPR framework, namely BEV$^2$PR, generating a composite descriptor with both visual cues and spatial awareness based on a single camera. The key points lie in: 1) We use BEV features as an explicit source of structural knowledge in constructing global features. 2) The lower layers of the pre-trained backbone from BEV generation are shared for visual and structural streams in VPR, facilitating the learning of fine-grained local features in the visual stream. 3) The complementary visual and structural features can jointly enhance VPR performance. Our BEV$^2$PR framework enables consistent performance improvements over several popular aggregation modules for RGB global features. The experiments on our collected VPR-NuScenes dataset demonstrate an absolute gain of 2.47% on Recall@1 for the strong Conv-AP baseline to achieve the best performance in our setting, and notably, a 18.06% gain on the hard set. The code and dataset will be available at https://github.com/FudongGe/BEV2PR.

Nian Liu, Jin Gao, Shubo Lin et al.

Infrared small target detection (ISTD) is challenging because tiny, low-contrast targets are easily obscured by complex and dynamic backgrounds. Conventional multi-frame approaches typically learn motion implicitly through deep neural networks, often requiring additional motion supervision or explicit alignment modules. We propose Motion Integration DETR (MI-DETR), a bio-inspired dual-pathway detector that processes one infrared frame per time step while explicitly modeling motion. First, a retina-inspired cellular automaton (RCA) converts raw frame sequences into a motion map defined on the same pixel grid as the appearance image, enabling parvocellular-like appearance and magnocellular-like motion pathways to be supervised by a single set of bounding boxes without extra motion labels or alignment operations. Second, a Parvocellular-Magnocellular Interconnection (PMI) Block facilitates bidirectional feature interaction between the two pathways, providing a biologically motivated intermediate interconnection mechanism. Finally, a RT-DETR decoder operates on features from the two pathways to produce detection results. Surprisingly, our proposed simple yet effective approach yields strong performance on three commonly used ISTD benchmarks. MI-DETR achieves 70.3% mAP@50 and 72.7% F1 on IRDST-H (+26.35 mAP@50 over the best multi-frame baseline), 98.0% mAP@50 on DAUB-R, and 88.3% mAP@50 on ITSDT-15K, demonstrating the effectiveness of biologically inspired motion-appearance integration. Code is available at https://github.com/nliu-25/MI-DETR.

Hanshi Wang, Yuhao Xu, Zekun Xu et al.

The established redundancy in visual tokens within large vision-language models allows pruning to effectively reduce their substantial computational demands. Previous methods typically employ heuristic layer-specific pruning strategies where, although the number of tokens removed may differ across decoder layers, the overall pruning schedule is fixed and applied uniformly to all input samples and tasks, failing to align token elimination with the model's holistic reasoning trajectory. Cognitive science indicates that human visual processing often begins with broad exploration to accumulate evidence before narrowing focus as the target becomes distinct. Our experiments reveal an analogous pattern in these models. This observation suggests that neither a fixed pruning schedule nor a heuristic layer-wise strategy can optimally accommodate the diverse complexities inherent in different inputs. To overcome this limitation, we introduce Complexity-Adaptive Pruning (AutoPrune), a training-free, plug-and-play framework that tailors pruning policies to varying sample and task complexities. Specifically, AutoPrune quantifies the mutual information between visual and textual tokens, then projects this signal to a budget-constrained logistic retention curve. Each such logistic curve, defined by its unique shape, corresponds to the specific complexity of different tasks and can guarantee adherence to predefined computational constraints. We evaluate AutoPrune on standard vision-language tasks and on Vision-Language-Action models for autonomous driving. Notably, when applied to LLaVA-1.5-7B, our method prunes 89% of visual tokens and reduces inference FLOPs by 76.8% while retaining 96.7% of the original accuracy averaged over all tasks. This corresponds to a 9.1% improvement over the recent work PDrop, demonstrating the effectiveness. Code is available at https://github.com/AutoLab-SAI-SJTU/AutoPrune.

Yifan Lu, Ziqi Zhang, Chunfeng Yuan et al.

Large Vision-Language Models (LVLMs) suffer from serious hallucination problems, where the model-generated responses are inconsistent with the visual inputs. Existing hallucination mitigation methods are mainly based on preference alignment and require external human annotations or auxiliary models for preference data collection, which increase costs and limit sustainable improvement. To tackle these challenges, we propose Autonomous Preference Alignment via Self-Injection (APASI), a novel and generalizable method that mitigates hallucinations without external dependencies. APASI leverages the target LVLM to self-inject hallucinations into a generated response, creating a pair of responses with varying preference levels. During the self-injection process, the dis-preferred response is generated based on three key observations of hallucinations, ensuring it simulates real hallucination patterns. This fidelity offers an accurate learning signal for hallucination mitigation. Moreover, APASI incorporates an iterative alignment training strategy combined with curriculum learning to periodically update the preference data with increasing challenge, enabling stable and continuous enhancement of the LVLM. Extensive experiments across six benchmarks show that APASI not only effectively mitigates hallucinations for three baseline models but also achieves comparable or even superior performance to alignment-based methods with external dependency, thereby demonstrating its effectiveness and generalization capability. The code is available at https://github.com/davidluciolu/APASI.

Sikui Zhang, Guangze Gao, Ziyun Gan et al.

Large language models (LLMs) experience significant performance degradation when the input exceeds the pretraining context window, primarily due to the out-of-distribution (OOD) behavior of Rotary Position Embedding (RoPE). Recent studies mitigate this problem by remapping OOD positions into the in-distribution range with fixed mapping strategies, ignoring the dynamic relationship between input length and the model's effective context window. To this end, we propose Length-aware Multi-grained Positional Encoding (LaMPE), a training-free method that fully utilizes the model's effective context window for adaptive long-context scaling in LLMs. Motivated by the left-skewed frequency distribution of relative positions, LaMPE establishes a dynamic relationship between mapping length and input length through a parametric scaled sigmoid function to adaptively allocate positional capacity across varying input lengths. Meanwhile, LaMPE devises a novel multi-grained attention mechanism that strategically allocates positional resolution across different sequence regions to capture both fine-grained locality and long-range dependencies. Our method can be seamlessly applied to a wide range of RoPE-based LLMs without training. Extensive experiments on three representative LLMs across five mainstream long-context benchmarks demonstrate that LaMPE achieves significant performance improvements compared to existing length extrapolation methods. The code will be released at https://github.com/scar-on/LaMPE.

Jiang Yuan, JI Ma, Bo Wang et al.

Implicit degradation estimation-based blind super-resolution (IDE-BSR) hinges on extracting the implicit degradation representation (IDR) of the LR image and adapting it to LR image features to guide HR detail restoration. Although IDE-BSR has shown potential in dealing with noise interference and complex degradations, existing methods ignore the importance of IDR discriminability for BSR and instead over-complicate the adaptation process to improve effect, resulting in a significant increase in the model's parameters and computations. In this paper, we focus on the discriminability optimization of IDR and propose a new powerful and lightweight BSR model termed LightBSR. Specifically, we employ a knowledge distillation-based learning framework. We first introduce a well-designed degradation-prior-constrained contrastive learning technique during teacher stage to make the model more focused on distinguishing different degradation types. Then we utilize a feature alignment technique to transfer the degradation-related knowledge acquired by the teacher to the student for practical inferencing. Extensive experiments demonstrate the effectiveness of IDR discriminability-driven BSR model design. The proposed LightBSR can achieve outstanding performance with minimal complexity across a range of blind SR tasks. Our code is accessible at: https://github.com/MJ-NCEPU/LightBSR.

Guanzhou Ke, Bo Wang, Guoqing Chao et al.

Multimodal foundation models have demonstrated impressive capabilities across diverse tasks. However, their potential as plug-and-play solutions for missing modality reconstruction remains underexplored. To bridge this gap, we identify and formalize three potential paradigms for missing modality reconstruction, and perform a comprehensive evaluation across these paradigms, covering 42 model variants in terms of reconstruction accuracy and adaptability to downstream tasks. Our analysis reveals that current foundation models often fall short in two critical aspects: (i) fine-grained semantic extraction from the available modalities, and (ii) robust validation of generated modalities. These limitations lead to suboptimal and, at times, misaligned generations. To address these challenges, we propose an agentic framework tailored for missing modality reconstruction. This framework dynamically formulates modality-aware mining strategies based on the input context, facilitating the extraction of richer and more discriminative semantic features. In addition, we introduce a self-refinement mechanism, which iteratively verifies and enhances the quality of generated modalities through internal feedback. Experimental results show that our method reduces FID for missing image reconstruction by at least 14\% and MER for missing text reconstruction by at least 10\% compared to baselines. Code are released at: https://github.com/Guanzhou-Ke/AFM2.

Yiwei Zhang, Jin Gao, Fudong Ge et al.

Bird's-eye-view (BEV) map layout estimation requires an accurate and full understanding of the semantics for the environmental elements around the ego car to make the results coherent and realistic. Due to the challenges posed by occlusion, unfavourable imaging conditions and low resolution, \emph{generating} the BEV semantic maps corresponding to corrupted or invalid areas in the perspective view (PV) is appealing very recently. \emph{The question is how to align the PV features with the generative models to facilitate the map estimation}. In this paper, we propose to utilize a generative model similar to the Vector Quantized-Variational AutoEncoder (VQ-VAE) to acquire prior knowledge for the high-level BEV semantics in the tokenized discrete space. Thanks to the obtained BEV tokens accompanied with a codebook embedding encapsulating the semantics for different BEV elements in the groundtruth maps, we are able to directly align the sparse backbone image features with the obtained BEV tokens from the discrete representation learning based on a specialized token decoder module, and finally generate high-quality BEV maps with the BEV codebook embedding serving as a bridge between PV and BEV. We evaluate the BEV map layout estimation performance of our model, termed VQ-Map, on both the nuScenes and Argoverse benchmarks, achieving 62.2/47.6 mean IoU for surround-view/monocular evaluation on nuScenes, as well as 73.4 IoU for monocular evaluation on Argoverse, which all set a new record for this map layout estimation task. The code and models are available on \url{https://github.com/Z1zyw/VQ-Map}.

Yice Zhang, Jie Zeng, Weiming Hu et al.

Aspect Sentiment Quad Prediction (ASQP) aims to predict all quads (aspect term, aspect category, opinion term, sentiment polarity) for a given review, which is the most representative and challenging task in aspect-based sentiment analysis. A key challenge in the ASQP task is the scarcity of labeled data, which limits the performance of existing methods. To tackle this issue, we propose a self-training framework with a pseudo-label scorer, wherein a scorer assesses the match between reviews and their pseudo-labels, aiming to filter out mismatches and thereby enhance the effectiveness of self-training. We highlight two critical aspects to ensure the scorer's effectiveness and reliability: the quality of the training dataset and its model architecture. To this end, we create a human-annotated comparison dataset and train a generative model on it using ranking-based objectives. Extensive experiments on public ASQP datasets reveal that using our scorer can greatly and consistently improve the effectiveness of self-training. Moreover, we explore the possibility of replacing humans with large language models for comparison dataset annotation, and experiments demonstrate its feasibility. We release our code and data at https://github.com/HITSZ-HLT/ST-w-Scorer-ABSA .

Yaya Shi, Xu Yang, Haiyang Xu et al.

Current metrics for video captioning are mostly based on the text-level comparison between reference and candidate captions. However, they have some insuperable drawbacks, e.g., they cannot handle videos without references, and they may result in biased evaluation due to the one-to-many nature of video-to-text and the neglect of visual relevance. From the human evaluator's viewpoint, a high-quality caption should be consistent with the provided video, but not necessarily be similar to the reference in literal or semantics. Inspired by human evaluation, we propose EMScore (Embedding Matching-based score), a novel reference-free metric for video captioning, which directly measures similarity between video and candidate captions. Benefit from the recent development of large-scale pre-training models, we exploit a well pre-trained vision-language model to extract visual and linguistic embeddings for computing EMScore. Specifically, EMScore combines matching scores of both coarse-grained (video and caption) and fine-grained (frames and words) levels, which takes the overall understanding and detailed characteristics of the video into account. Furthermore, considering the potential information gain, EMScore can be flexibly extended to the conditions where human-labeled references are available. Last but not least, we collect VATEX-EVAL and ActivityNet-FOIl datasets to systematically evaluate the existing metrics. VATEX-EVAL experiments demonstrate that EMScore has higher human correlation and lower reference dependency. ActivityNet-FOIL experiment verifies that EMScore can effectively identify "hallucinating" captions. The datasets will be released to facilitate the development of video captioning metrics. The code is available at: https://github.com/ShiYaya/emscore.

Zhipeng Zhang, Yihao Liu, Xiao Wang et al.

Siamese tracking has achieved groundbreaking performance in recent years, where the essence is the efficient matching operator cross-correlation and its variants. Besides the remarkable success, it is important to note that the heuristic matching network design relies heavily on expert experience. Moreover, we experimentally find that one sole matching operator is difficult to guarantee stable tracking in all challenging environments. Thus, in this work, we introduce six novel matching operators from the perspective of feature fusion instead of explicit similarity learning, namely Concatenation, Pointwise-Addition, Pairwise-Relation, FiLM, Simple-Transformer and Transductive-Guidance, to explore more feasibility on matching operator selection. The analyses reveal these operators' selective adaptability on different environment degradation types, which inspires us to combine them to explore complementary features. To this end, we propose binary channel manipulation (BCM) to search for the optimal combination of these operators. BCM determines to retrain or discard one operator by learning its contribution to other tracking steps. By inserting the learned matching networks to a strong baseline tracker Ocean, our model achieves favorable gains by $67.2 \rightarrow 71.4$, $52.6 \rightarrow 58.3$, $70.3 \rightarrow 76.0$ success on OTB100, LaSOT, and TrackingNet, respectively. Notably, Our tracker, dubbed AutoMatch, uses less than half of training data/time than the baseline tracker, and runs at 50 FPS using PyTorch. Code and model will be released at https://github.com/JudasDie/SOTS.

Li Yang, Yan Xu, Shaoru Wang et al.

Recent one-stage object detectors follow a per-pixel prediction approach that predicts both the object category scores and boundary positions from every single grid location. However, the most suitable positions for inferring different targets, i.e., the object category and boundaries, are generally different. Predicting all these targets from the same grid location thus may lead to sub-optimal results. In this paper, we analyze the suitable inference positions for object category and boundaries, and propose a prediction-target-decoupled detector named PDNet to establish a more flexible detection paradigm. Our PDNet with the prediction decoupling mechanism encodes different targets separately in different locations. A learnable prediction collection module is devised with two sets of dynamic points, i.e., dynamic boundary points and semantic points, to collect and aggregate the predictions from the favorable regions for localization and classification. We adopt a two-step strategy to learn these dynamic point positions, where the prior positions are estimated for different targets first, and the network further predicts residual offsets to the positions with better perceptions of the object properties. Extensive experiments on the MS COCO benchmark demonstrate the effectiveness and efficiency of our method. With a single ResNeXt-64x4d-101-DCN as the backbone, our detector achieves 50.1 AP with single-scale testing, which outperforms the state-of-the-art methods by an appreciable margin under the same experimental settings.Moreover, our detector is highly efficient as a one-stage framework. Our code is public at https://github.com/yangli18/PDNet.

Chao Liang, Zhipeng Zhang, Xue Zhou et al.

The one-shot multi-object tracking, which integrates object detection and ID embedding extraction into a unified network, has achieved groundbreaking results in recent years. However, current one-shot trackers solely rely on single-frame detections to predict candidate bounding boxes, which may be unreliable when facing disastrous visual degradation, e.g., motion blur, occlusions. Once a target bounding box is mistakenly classified as background by the detector, the temporal consistency of its corresponding tracklet will be no longer maintained. In this paper, we set out to restore the bounding boxes misclassified as ``fake background'' by proposing a re-check network. The re-check network innovatively expands the role of ID embedding from data association to motion forecasting by effectively propagating previous tracklets to the current frame with a small overhead. Note that the propagation results are yielded by an independent and efficient embedding search, preventing the model from over-relying on detection results. Eventually, it helps to reload the ``fake background'' and repair the broken tracklets. Building on a strong baseline CSTrack, we construct a new one-shot tracker and achieve favorable gains by 70.7 $\rightarrow$ 76.4, 70.6 $\rightarrow$ 76.3 MOTA on MOT16 and MOT17, respectively. It also reaches a new state-of-the-art MOTA and IDF1 performance. Code is released at https://github.com/JudasDie/SOTS.