多模态大模型

2512.12107 2025-12-16 cs.CV 83%

EchoVLM: Measurement-Grounded Multimodal Learning for Echocardiography

EchoVLM：基于测量的多模态学习用于超声心动图

Yuheng Li, Yue Zhang, Abdoul Aziz Amadou, Yuxiang Lai, Jike Zhong, Tiziano Passerini, Dorin Comaniciu, Puneet Sharma

机构 * Georgia Institute of Technology（佐治亚理工学院）； Siemens Healthineers（西门子医疗）； Siemens Healthcare Limited（西门子医疗有限公司）； Emory University（埃默里大学）； University of Southern California（南加州大学）

专题命中图文多模态：multimodal(title,abstract)；image-text(abstract)；分类 cs.CV

AI总结 EchoVLM通过引入基于测量的多模态学习方法，实现了超声心动图的端到端解读，提升了疾病分类和视图识别的性能。

2509.23661 2025-12-16 cs.CV 79%

LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training

LLaVA-OneVision-1.5：面向民主化多模态训练的完全开放框架

Xiang An, Yin Xie, Kaicheng Yang, Wenkang Zhang, Xiuwei Zhao, Zheng Cheng, Yirui Wang, Songcen Xu, Changrui Chen, Didi Zhu, Chunsheng Wu, Huajie Tan, Chunyuan Li, Jing Yang, Jie Yu, Xiyao Wang, Bin Qin, Yumeng Wang, Zizhen Yan, Ziyong Feng, Ziwei Liu, Bo Li, Jiankang Deng

机构 * LLaVA-OneVision Community Contributors（LLaVA-OneVision社区贡献者）

专题命中图文多模态：multimodal(title,abstract)；分类 cs.CV

AI总结 LLaVA-OneVision-1.5通过开放框架和高效训练方法实现了多模态模型的低成本高性能训练。

Comments LLaVA-OneVision-1.5 Technical Report

2505.23065 2025-12-16 cs.CL 79%

SNS-Bench-VL: Benchmarking Multimodal Large Language Models in Social Networking Services

SNS-Bench-VL：在社交网络服务中评估多模态大语言模型的基准测试

Hongcheng Guo, Zheyong Xie, Shaosheng Cao, Boyang Wang, Weiting Liu, Anjie Le, Lei Li, Zhoujun Li

专题命中图文多模态：multimodal(title,abstract)；分类 cs.CL

AI总结 SNS-Bench-VL是一个用于评估多模态大语言模型在社交网络服务中性能的基准测试，涵盖8个多模态任务，包含4001个问题-答案对，评估25种先进模型，揭示多模态社交理解的挑战。

Comments We found problems in the code while rechecking our implementation. These issues led to noticeable numerical discrepancies, making some of the reported results and conclusions potentially unreliable. Therefore, we request to withdraw this submission

2512.12701 2025-12-16 cs.CV cs.CL cs.LG 62%

Efficient Vision-Language Reasoning via Adaptive Token Pruning

通过自适应令牌修剪实现高效的视觉语言推理

Xue Li, Xiaonan Song, Henry Hu

机构 * Scholar42（学者42）； InfiniPouch LLC（InfiniPouch公司）； Labelbox, Inc.（Labelbox公司）

专题命中图文多模态：multimodal(abstract)；分类 cs.CV、cs.CL

AI总结本研究提出自适应令牌修剪方法，通过动态保留关键令牌提升视觉语言模型的推理效率和稳定性，减少计算量并保持精度。

Comments 10 pages, 3 figures. Expanded version of an extended abstract accepted at NeurIPS 2025 Workshop on VLM4RWD. Presents methodology and preliminary experimental results

2512.12487 2025-12-16 cs.CV 57%

Relation-R1: Progressively Cognitive Chain-of-Thought Guided Reinforcement Learning for Unified Relation Comprehension

Relation-R1: 逐步认知链式思维引导的强化学习用于统一关系理解

Lin Li, Wei Chen, Jiahui Li, Kwang-Ting Cheng, Long Chen

专题命中图文多模态：multi-modal(abstract)；分类 cs.CV

AI总结 Relation-R1通过整合认知链式思维引导的强化学习，实现了对二元和N元关系的统一理解，提升了视觉-语义定位能力。

Comments AAAI 2026

1. 图文多模态 6 篇

EchoVLM: Measurement-Grounded Multimodal Learning for Echocardiography

LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training

SNS-Bench-VL: Benchmarking Multimodal Large Language Models in Social Networking Services

Efficient Vision-Language Reasoning via Adaptive Token Pruning

More Than the Final Answer: Improving Visual Extraction and Logical Consistency in Vision-Language Models

Relation-R1: Progressively Cognitive Chain-of-Thought Guided Reinforcement Learning for Unified Relation Comprehension