Large Vision Language Model
Large Vision-Language Models (LVLMs) integrate computer vision and natural language processing to enable machines to understand and reason about images and text simultaneously. Current research focuses on improving LVLMs' accuracy, efficiency, and robustness, particularly addressing issues like hallucinations (generating inaccurate information), and enhancing their ability to perform multi-level visual perception and reasoning tasks, including quantitative spatial reasoning and mechanical understanding. These advancements are significant for various applications, including medical image analysis, robotics, and autonomous driving, by enabling more reliable and insightful multimodal data processing.
Papers
VideoLLM-MoD: Efficient Video-Language Streaming with Mixture-of-Depths Vision Computation
Shiwei Wu, Joya Chen, Kevin Qinghong Lin, Qimeng Wang, Yan Gao, Qianli Xu, Tong Xu, Yao Hu, Enhong Chen, Mike Zheng Shou
LLaVA-SG: Leveraging Scene Graphs as Visual Semantic Expression in Vision-Language Models
Jingyi Wang, Jianzhong Ju, Jian Luan, Zhidong Deng
CluMo: Cluster-based Modality Fusion Prompt for Continual Learning in Visual Question Answering
Yuliang Cai, Mohammad Rostami
Swarm Intelligence in Geo-Localization: A Multi-Agent Large Vision-Language Model Collaborative Framework
Xiao Han, Chen Zhu, Xiangyu Zhao, Hengshu Zhu
Making Large Vision Language Models to be Good Few-shot Learners
Fan Liu, Wenwen Cai, Jian Huo, Chuanyi Zhang, Delong Chen, Jun Zhou
Towards Analyzing and Mitigating Sycophancy in Large Vision-Language Models
Yunpu Zhao, Rui Zhang, Junbin Xiao, Changxin Ke, Ruibo Hou, Yifan Hao, Qi Guo, Yunji Chen