Vision Model
Vision models are artificial intelligence systems designed to interpret and understand visual information, aiming to replicate aspects of human visual perception and reasoning. Current research emphasizes improving efficiency and generalization across diverse tasks, focusing on architectures like Vision Transformers and Convolutional Neural Networks, often incorporating large language models for multimodal understanding and instruction following. This field is crucial for advancing various applications, from medical image analysis and robotic manipulation to enhancing accessibility and creative tools, with ongoing efforts to improve model robustness, explainability, and alignment with human perception.
Papers
Visual Program Distillation: Distilling Tools and Programmatic Reasoning into Vision-Language Models
Yushi Hu, Otilia Stretcu, Chun-Ta Lu, Krishnamurthy Viswanathan, Kenji Hata, Enming Luo, Ranjay Krishna, Ariel Fuxman
Machine Vision Therapy: Multimodal Large Language Models Can Enhance Visual Robustness via Denoising In-Context Learning
Zhuo Huang, Chang Liu, Yinpeng Dong, Hang Su, Shibao Zheng, Tongliang Liu