Weiming Ren

Offline multi-agent reinforcement learning for evaluating and optimizing football attacking strategies against low-block defences

Yi Pan, Zhiqiang Pu, Min Chen, Weiming Ren et autres

Applying reinforcement learning to real-world multi-agent systems remains challenging due to coordination complexity, partial observability, and the constraints of learning from static datasets. Football serves as a premier and fascinating testbed for these issues, as exemplified by the ‘low block’, a representative …

cn (code pays fourni par la source)

TUNA: Taming Unified Visual Representations for Native Unified Multimodal Models

Zhiheng Liu, Weiming Ren, Zijian Zhou, Shoufa Chen et autres

Unified multimodal models (UMMs) aim to jointly perform multimodal understanding and generation within a single framework. We present TUNA, a native UMM that builds a unified continuous visual representation by cascading a VAE encoder with a representation encoder. This unified representation space …

Network pharmacology-based analysis of Cornus officinalis for anti-osteoporosis property and experimental validation of its active component beta-sitosterol

Xiaoming Gao, Chao Lin, Weiming Ren, Yanming Hao et autres

Introduction Cornus officinalis is a commonly used herbal for osteoporosis. We aimed to explore the potential mechanism of Cornus officinalis in the treatment of postmenopausal osteoporosis using network pharmacology. Materials and Methods The active ingredients and targets of Cornus officinalis were screened …

cn (code pays fourni par la source)

VAMBA: Understanding Hour-Long Videos with Hybrid Mamba-Transformers

Weiming Ren, Wentao Ma, Huan Yang, Ge Zhang et autres

State-of-the-art transformer-based large multimodal models (LMMs) struggle to handle hour-long video inputs due to the quadratic complexity of the causal self-attention operations, leading to high computational costs during training and inference. Existing token compression-based methods reduce the number of video tokens but …

ca (code pays fourni par la source)

Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learning

Haozhe Wang, A.W.Y. Su, Weiming Ren, Fangzhen Lin et autres

Chain-of-thought reasoning has significantly improved the performance of Large Language Models (LLMs) across various domains. However, this reasoning process has been confined exclusively to textual space, limiting its effectiveness in visually intensive tasks. To address this limitation, we introduce the concept of …

VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation

Wentao Ma, Weiming Ren, Yiming Jia, Zhuofeng Li et autres

Large multimodal models (LMMs) have recently emerged as a powerful tool for long video understanding (LVU), prompting the development of standardized LVU benchmarks to evaluate their performance. However, our investigation reveals a rather sober lesson for existing LVU benchmarks. First, most existing …

Frequency-Enhanced Dual-Domain Attention Network for Automotive Intrusion Detection System

Weiming Ren, Yongyi Chen, Dan Zhang

For the security of autonomous vehicles in the Internet of Vehicles (IoV), the intrusion detection system (IDS) is developed to detect malicious behavior and identify potential threats. Deep learning (DL)-based IDS for controller area network (CAN) bus have shown excellent performance. However, …

cn (code pays fourni par la source)

Determination of Cyazofamid and Its Metabolite in Rice Based on Magnetic Zirconia Nanoparticles Sample Cleanup and Its Application to a Dissipation and Residue Study

Dan Zhao, Jianguo Liu, Yiqi Yang, Jiaxin Zhang et autres

An analytical method was developed for the determination of cyazofamid (CZFM) and its metabolite CCIM in rice, employing magnetic zirconia nanoparticles (MZNPs) for sample cleanup. MZNPs were synthesized through a one-step hydrothermal process and characterized by multiple techniques. Samples of rice plant, …

cn (code pays fourni par la source)

OmniEdit: Building Image Editing Generalist Models Through Specialist Supervision

Cong Wei, Zhiqiang Xiong, Weiming Ren, Xinrun Du et autres

Instruction-guided image editing methods have demonstrated significant potential by training diffusion models on automatically synthesized or manually annotated image editing pairs. However, these methods remain far from practical, real-life applications. We identify three primary challenges contributing to this gap. Firstly, existing models …

Bi-level weighted mixed-domain self-attention network for non-contact heart rate estimation

Weiming Ren, Yongyi Chen, Dan Zhang, Hamid Reza Karimi

cn, it (code pays fourni par la source)

MMMU: A Massive Multi-Discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI

Yue Xiang, Yuansheng Ni, Tianyu Zheng, Kai Zhang et autres

We introduce MMMU: a new benchmark designed to evaluate multimodal models on massive multi-discipline tasks demanding college-level subject knowledge and deliberate reasoning. MMMU includes 11.5K meticulously collected multimodal questions from college exams, quizzes, and text-books, covering six core disciplines: Art & Design, …

ca, us (code pays fourni par la source)

Aqueous degradation of tioxazafen: Kinetics, products, pathways, and toxicity

Zhou Lü, Fuying Zhang, Yiqi Yang, Hong Zhang et autres

cn (code pays fourni par la source)