Current visual representation learning remains bifurcated: vision-language models (e.g., CLIP) excel at global semantic alignment but lack spatial precision, while self-supervised methods (e.g., MAE, DINO) capture intricate local structures yet struggle with high-level semantic context. We argue that these paradigms are fundamentally complementary and can be integrated into a principled multi-task framework, further enhanced by dense spatial supervision. We introduce MTV, a multi-task visual pretraining framework that jointly optimizes a shared backbone across vision-language contrastive, self-supervised, and dense spatial objectives. To mitigate the need for manual annotations, we leverage high-capacity "expert" models--such as Depth Anything V2 and OWLv2--to synthesize dense, structured pseudo-labels at scale. Beyond the framework, we provide a systematic investigation into the mechanics of multi-task visual learning, analyzing: (i) the individual and marginal gain of each objective, (ii) task synergies versus interference, and (iii) scaling behavior across varying data and model scales. Our results demonstrate that MTV achieves "best-of-both-worlds" performance, significantly enhancing fine-grained spatial reasoning without compromising global semantic understanding. Our findings suggest that multi-task learning, fueled by high-quality pseudo-supervision, is a scalable path toward more general visual encoders.
Position Rebinding Cache Reuse: Replay-Free Visual Revisiting for Interleaved Multimodal Reasoning
Sun Yat-sen University（中山大学）；Shenzhen Loop Area Institute（深圳河套学院）；Harbin Institute of Technology (HIT)（哈尔滨工业大学）；Fudan University（复旦大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；The Chinese University of Hong Kong（香港中文大学）
Portland State University（波特兰州立大学）；Stanford University（斯坦福大学）；Santa Fe Institute（圣塔菲研究所）
We revisit a particular visual grounding method: the "Image Retrieval Using Scene Graphs" (IRSG) system of Johnson et al. (2015). Our experiments indicate that the system does not effectively use its learned object-relationship models. We also look closely at the IRSG dataset, as well as the widely used Visual Relationship Dataset (VRD) that is adapted from it. We find that these datasets exhibit biases that allow methods that ignore relationships to perform relatively well. We also describe several other problems with the IRSG dataset, and report on experiments using a subset of the dataset i…
This paper focuses on visual counting, which aims to predict the number of occurrences given a natural image and a query (e.g. a question or a category). Unlike most prior works that use explicit, symbolic models which can be computationally expensive and limited in generalization, we propose a simple and effective alternative by revisiting modulated convolutions that fuse the query and the image locally. Following the design of residual bottleneck, we call our method MoVie, short for Modulated conVolutional bottlenecks. Notably, MoVie reasons implicitly and holistically and only needs a singl…
Visual question answering (VQA) is an interesting learning setting for evaluating the abilities and shortcomings of current systems for image understanding. Many of the recently proposed VQA systems include attention or memory mechanisms designed to support "reasoning". For multiple-choice VQA, nearly all of these systems train a multi-class classifier on image and question features to predict an answer. This paper questions the value of these common practices and develops a simple alternative model based on binary classification. Instead of treating answers as competing choices, our model rec…
Don't Just Chase "Highlighted Tokens" in MLLMs: Revisiting Visual Holistic Context Retention
The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州））；The Hong Kong University of Science and Technology（香港科学与技术大学）；Shanghai Jiao Tong University（上海交通大学）
Shandong University（山东大学）；Harbin Institute of Technology, Shenzhen（哈尔滨工业大学（深圳））
Recent studies have pointed out that many well-developed Visual Question Answering (VQA) models are heavily affected by the language prior problem, which refers to making predictions based on the co-occurrence pattern between textual questions and answers instead of reasoning visual contents. To tackle it, most existing methods focus on enhancing visual feature learning to reduce this superficial textual shortcut influence on VQA model decisions. However, limited effort has been devoted to providing an explicit interpretation for its inherent cause. It thus lacks a good guidance for the resear…
Microsoft Corporation（微软公司）；University of Washington（华盛顿大学）
This paper presents a detailed study of improving visual representations for vision language (VL) tasks and develops an improved object detection model to provide object-centric representations of images. Compared to the most widely used \emph{bottom-up and top-down} model \cite{anderson2018bottom}, the new model is bigger, better-designed for VL tasks, and pre-trained on much larger training corpora that combine multiple public annotated object detection datasets. Therefore, it can generate representations of a richer collection of visual objects and concepts. While previous VL research focus…
Revisiting Visual Understanding in Multimodal Reasoning through a Lens of Image Perturbation
School of Computer Science, Shanghai Jiao Tong University（上海交通大学计算机科学学院）；Shanghai Innovation Institute（上海创新研究院）；Zhongguancun Academy（中关村学院）；Tencent（腾讯）；State Key Laboratory of General Artificial Intelligence, BIGAI（通用人工智能国家重点实验室，BIGAI）；Lehigh University（莱斯大学）
SSL4RL: Revisiting Self-supervised Learning as Intrinsic Reward for Visual-Language Reasoning
Peking University（北京大学）；MIT（麻省理工学院）；TUM（技术大学（TUM））；Meituan（美团）；Peking University School of CIT, MCML, MDSIEECS and CSAIL（北京大学计算机学院、MCML、MDSIEECS以及CSAIL）
Revisiting CroPA: A Reproducibility Study and Enhancements for Cross-Prompt Adversarial Transferability in Vision-Language Models
Mehta Family School of Data Science and Artificial Intelligence（梅hta家族数据科学与人工智能学院）；Indian Institute of Technology, Roorkee（印度理工学院罗奥克学院）；Department of Civil Engineering（土木工程系）；Department of Electronics and Communication（电子与通信系）
From Pixels to Tokens: Revisiting Object Hallucinations in Large Vision-Language Models
University of Chinese Academy of Sciences（中国科学院大学）；Dept. of Comp. Sci. and Tech., Institute for AI, Tsinghua University（计算机科学与技术系，人工智能研究院，清华大学）；Gaoling School of Artificial Intelligence, Renmin University of China（人工智能学院，中国人民大学）；Kuaishou Technology Inc.（快手科技有限公司）；Shanghai Key Laboratory of Multi. Info. Processing, East China Normal University（多信息处理重点实验室，华东师范大学）
Revisiting Change VQA in Remote Sensing with Structured and Native Multimodal Qwen Models
Computer Engineering Department, College of Computer and Information Sciences, King Saud University（计算机工程系，计算机与信息科学学院，沙特国王大学）；Applied Computer Science Department, College of Applied Computer Science, King Saud University（应用计算机科学系，应用计算机科学学院，沙特国王大学）
Vision-and-Language Navigation (VLN) is a task that an agent is required to follow a language instruction to navigate to the goal position, which relies on the ongoing interactions with the environment during moving. Recent Transformer-based VLN methods have made great progress benefiting from the direct connections between visual observations and the language instruction via the multimodal cross-attention mechanism. However, these methods usually represent temporal context as a fixed-length vector by using an LSTM decoder or using manually designed hidden states to build a recurrent Transform…
The University of Sydney（悉尼大学）；Baidu Inc.（百度公司）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）
Transferring knowledge from task-agnostic pre-trained deep models for downstream tasks is an important topic in computer vision research. Along with the growth of computational capacity, we now have open-source vision-language pre-trained models in large scales of the model architecture and amount of data. In this study, we focus on transferring knowledge for video classification tasks. Conventional methods randomly initialize the linear classifier head for vision classification, but they leave the usage of the text encoder for downstream visual recognition tasks undiscovered. In this paper, w…
LaV-CoT: Language-Aware Visual CoT with Multi-Aspect Reward Optimization for Real-World Multilingual VQA
Ant Digital Technologies, Ant Group（蚂蚁集团数字技术部）；College of Computing and Data Science, Nanyang Technological University（南洋理工大学计算机与数据科学学院）；CFAR, Agency for Science, Technology and Research（科技研究局CFAR）
Revisiting Backdoor Attacks against Large Vision-Language Models from Domain Shift
National University of Singapore（新加坡国立大学）；Shenzhen Campus of Sun Yat-sen University（中山大学深圳校区）；Sea AI lab（海天人工智能实验室）；The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））；Nanyang Technological University（南洋理工大学）
Audio-Sync Video Generation with Multi-Stream Temporal Control
Beijing Academy of Artificial Intelligence（北京人工智能研究院）；School of Software and Microelectronics, Peking University（北京大学软件与微电子学院）；School of Artificial Intelligence, Beijing University of Posts and Telecommunications（北京邮电大学人工智能学院）；Nat’l Key Lab of General AI, School of Intelligence Science and Technology, Peking University（国家通用人工智能实验室，北京大学智能科学与技术学院）；Nat’l Eng. Research Ctr. of Visual Tech., School of Computer Science, Peking University（国家视觉技术工程研究中心，北京大学计算机学院）
本文提出 MTV 框架，通过分离语音、音效和音乐多流音频实现细粒度音视频同步控制，并构建 DEMIX 数据集，在六项指标上达到最先进性能。
Revisiting KRISP: A Lightweight Reproduction and Analysis of Knowledge-Enhanced Vision-Language Models
Centre For Digital Health（数字健康中心）；Indian Institute of Technology Bombay（孟买印度理工学院）；Department of Metallurgical engineering & Material science（冶金工程与材料科学系）；Department of Bioscience & Bioengineering（生物科学与生物工程系）
Visual tracking algorithms are naturally adopted in various applications, there have been several benchmarks and many tracking algorithms, more expected to appear in the future. In this report, I focus on single object tracking and revisit the details of tracker evaluation based on widely used OTB\cite{otb} benchmark by introducing a simpler, accurate, and extensible method for tracker evaluation and comparison. Experimental results suggest that there may not be an absolute winner among tracking algorithms. We have to perform detailed analysis to select suitable trackers for use cases.
Beyond Templates: Revisiting Zero-Shot Remote Sensing through Meta-Prompting
Remote Sensing Lab, National Technical University of Athens（遥感实验室，国家技术大学雅典分校）；Remote Sensing Lab, Department of Engineering and Sciences, National Technical University of Athens（遥感实验室，工程与科学系，国家技术大学雅典分校）；Universitas Mercatorum（默卡托大学）
Unsupervised visual representation learning remains a largely unsolved problem in computer vision research. Among a big body of recently proposed approaches for unsupervised learning of visual representations, a class of self-supervised techniques achieves superior performance on many challenging benchmarks. A large number of the pretext tasks for self-supervised learning have been studied, but other important aspects, such as the choice of convolutional neural networks (CNN), has not received equal attention. Therefore, we revisit numerous previously proposed self-supervised models, conduct a…
Revisiting Multimodal KV Cache Compression: A Frequency-Domain-Guided Outlier-KV-Aware Approach
College of Future Information Technology, Fudan University（未来信息科技学院，复旦大学）；Shanghai Innovation Institute（上海创新研究院）；The Chinese University of Hong Kong（香港中文大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Zhangjiang Laboratory（张江实验室）
Model pre-training is a cornerstone of modern visual recognition systems. Although fully supervised pre-training on datasets like ImageNet is still the de-facto standard, recent studies suggest that large-scale weakly supervised pre-training can outperform fully supervised approaches. This paper revisits weakly-supervised pre-training of models using hashtag supervision with modern versions of residual networks and the largest-ever dataset of images and corresponding hashtags. We study the performance of the resulting models in various transfer-learning settings including zero-shot transfer. W…
Grounding Emotion Recognition with Visual Prototypes: VEGA -- Revisiting CLIP in MERC
Xi'an Jiaotong University（西安交通大学）；Queen Mary University of London（伦敦女王玛丽大学）；Center for Multimodal AI（多模态人工智能中心）；Digital Environment Research Institute（数字环境研究院）
The recent advances in 3D sensing technology have made possible the capture of point clouds in significantly high resolution. However, increased detail usually comes at the expense of high storage, as well as computational costs in terms of processing and visualization operations. Mesh and Point Cloud simplification methods aim to reduce the complexity of 3D models while retaining visual quality and relevant salient features. Traditional simplification techniques usually rely on solving a time-consuming optimization problem, hence they are impractical for large-scale datasets. In an attempt to…
Contrastive learning-based vision-language pre-training approaches, such as CLIP, have demonstrated great success in many vision-language tasks. These methods achieve cross-modal alignment by encoding a matched image-text pair with similar feature embeddings, which are generated by aggregating information from visual patches and language tokens. However, direct aligning cross-modal information using such representations is challenging, as visual patches and text tokens differ in semantic levels and granularities. To alleviate this issue, we propose a Finite Discrete Tokens (FDT) based multimod…
Revisiting Shadow Detection from a Vision-Language Perspective
CAS Key Laboratory of Technology in Geo-spatial Information Processing and Application System, Department of Electronic Engineering and Information Science, University of Science and Technology of China（中国科学技术大学地球空间信息处理与应用系统重点实验室，电子工程与信息科学系）
The trend towards increasingly deep neural networks has been driven by a general observation that increasing depth increases the performance of a network. Recently, however, evidence has been amassing that simply increasing depth may not be the best way to increase performance, particularly given other limitations. Investigations into deep residual networks have also suggested that they may not in fact be operating as a single deep network, but rather as an ensemble of many relatively shallow networks. We examine these issues, and in doing so arrive at a new interpretation of the unravelled vi…
