Bin Li

Video anomaly detection (VAD) has been widely studied for its important applications in multimedia community. Recently, many Weakly Supervised VAD (WS-VAD) methods have been proposed, which tend to treat VAD as a classification task through multiple instance learning and result in the need to collect sufficient anomaly classes and samples to be used for training a classifier. However, anomaly events tend to be open-set and rare in real-world applications, so we often have difficulty collecting all anomaly classes and enough sample anomalies, which is a difficult situation for WS-VAD to cope with. To this end, we consider to treat VAD as an out-of-distribution detection task rather than a classification task and propose a simple but effective semi-supervised baseline method. First, we leverage the powerful zero-shot capability of large visual language models to generate summary text descriptions for videos and extract visual features as intermediates for subsequent use. Next, we use a text encoder to extract language features and combine them with visual features to obtain robust multimodal features. Finally, we introduce an out-of-distribution detection method that learns the center of normality in multimodal space from normal and unlabeled samples, while deviating abnormal samples from the center to cope with the scarcity of abnormal samples. To implement our baseline method, we also provide a new semi-supervised dataset by reorganizing an existing benchmark, which is the first available dataset in the VAD community that provides trimmed videos consisting of complete abnormal events. Experiments demonstrate that our method performs more robustly when fewer anomaly classes and anomaly samples collected.

Large Vision-Language Models (LVLMs) have recently demonstrated strong multimodal understanding, yet their fine-grained visual perception is often constrained by low input resolutions. A common remedy is to partition high-resolution images into multiple sub-images for separate encoding, but this approach drastically inflates the number of visual tokens and introduces prohibitive inference overhead. To overcome this challenge, we propose Pyramid Token Pruning (PTP), a training-free strategy that hierarchically integrates bottom-up visual saliency at both region and token levels with top-down instruction-guided relevance. Inspired by human visual cognition, PTP selectively preserves more tokens from salient regions while further emphasizing those most relevant to task instructions. Extensive experiments on 13 diverse benchmarks show that PTP substantially reduces computational cost, memory usage, and inference latency, with negligible performance degradation.

The abstract visual reasoning ability in human intelligence benefits discovering underlying rules in the novel environment. Raven’s Progressive Matrix (RPM) is a classic test to realize such ability in machine intelligence by selecting from candidates. Recent studies suggest that solving RPM in an answer-generation way boosts a more in-depth understanding of rules. However, existing generative solvers cannot discover the global concept-changing rules without auxiliary supervision (e.g., rule annotations and distractors in candidate sets). To this end, we propose a deep latent variable model for Concept-changing Rule ABstraction (CRAB) by learning interpretable concepts and parsing concept-changing rules in the latent space. With the iterative learning process, CRAB can automatically abstract global rules shared on the dataset on each concept and form the learnable prior knowledge of global rules. CRAB outperforms the baselines trained without auxiliary supervision in the arbitrary-position answer generation task and achieves comparable and even higher accuracy than the compared models trained with auxiliary supervision. Finally, we conduct experiments to illustrate the interpretability of CRAB in concept learning, answer selection, and global rule abstraction.

Computer-Aided Design (CAD) is essential in industrial design, but the complexity of traditional CAD modeling and workflows presents significant challenges for automating the generation of high-precision, editable CAD models. Existing methods that reconstruct 3D models from sketches often produce non-editable and approximate models that fall short of meeting the stringent requirements for precision and editability in industrial design. Moreover, the reliance on text or image-based inputs often requires significant manual annotation, limiting their scalability and applicability in industrial settings. To overcome these challenges, we propose the Heterogeneous Collaborative Multi-Expert Reinforcement Learning (CME-CAD) paradigm, a novel training paradigm for CAD code generation. Our approach integrates the complementary strengths of these models, facilitating collaborative learning and improving the model's ability to generate accurate, constraint-compatible, and fully editable CAD models. We introduce a two-stage training process: Multi-Expert Fine-Tuning (MEFT), and Multi-Expert Reinforcement Learning (MERL). Additionally, we present CADExpert, an open-source benchmark consisting of 17,299 instances, including orthographic projections with precise dimension annotations, expert-generated Chain-of-Thought (CoT) processes, executable CADQuery code, and rendered 3D models.

With rich temporal-spatial information, video-based person re-identification methods have shown broad prospects. Although tracklets can be easily obtained with ready-made tracking models, annotating identities is still expensive and impractical. Therefore, some video-based methods propose using only a few identity annotations or camera labels to facilitate feature learning. They also simply average the frame features of each tracklet, overlooking unexpected variations and inherent identity consistency within tracklets. In this paper, we propose the Self-Supervised Refined Clustering (SSR-C) framework without relying on any annotation or auxiliary information to promote unsupervised video person re-identification. Specifically, we first propose the Noise-Filtered Tracklet Partition (NFTP) module to reduce the feature bias of tracklets caused by noisy tracking results, and sequentially partition the noise-filtered tracklets into “sub-tracklets”. Then, we cluster and further merge sub-tracklets using the self-supervised signal from the tracklet partition, which is enhanced through a progressive strategy to generate reliable pseudo labels, facilitating intra-class cross-tracklet aggregation. Moreover, we propose the Class Smoothing Classification (CSC) loss to efficiently promote model learning. Extensive experiments on the MARS and DukeMTMC-VideoReID datasets demonstrate that our proposed SSR-C for unsupervised video person re-identification achieves state-of-the-art results and is comparable to advanced supervised methods.

LVLMs have been shown to perform excellently in image-level tasks such as VQA and caption. However, in many instance-level tasks, such as visual grounding and object detection, LVLMs still show performance gaps compared to previous expert models. Meanwhile, although pedestrian tracking is a classical task, there have been a number of new topics in combining object tracking and natural language, such as Referring MOT, Cross-view Referring MOT, and Semantic MOT. These tasks emphasize that models should understand the tracked object at an advanced semantic level, which is exactly where LVLMs excel. In this paper, we propose a new unified Pedestrian Tracking framework, namely OmniPT, which can track, track based on reference and generate semantic understanding of tracked objects interactively. We address two issues: how to model the tracking task into a task that foundation models can perform, and how to make the model output formatted answers. To this end, we implement a training phase consisting of RL-Mid Training-SFT-RL. Based on the pre-trained weights of the LVLM, we first perform a simple RL phase to enable the model to output fixed and supervisable bounding box format. Subsequently, we conduct a mid-training phase using a large number of pedestrian-related datasets. Finally, we perform supervised fine-tuning on several pedestrian tracking datasets, and then carry out another RL phase to improve the model's tracking performance and enhance its ability to follow instructions. We conduct experiments on tracking benchmarks and the experimental results demonstrate that the proposed method can perform better than the previous methods.