达摩院联合浙大、港理工推出PixelRefer：多模态大模型迈向像素级视觉理解
"Vision-Language Models" 相关文章
Dynamic Routing Between Experts: A Data-Efficient Approach to Continual Learning in Vision-Language Models
世界模型==VQA？机器人不用想象画面，预测语义就够了
Token-Level Inference-Time Alignment for Vision-Language Models
世界模型==VQA？机器人不用想象画面，预测语义就够了
不重构、不牺牲通用性：VLM-FO1，为任何VLM无损增强细粒度感知能力
DeepSeek的终极野心：把大语言模型的基本语言都改造成图像
MIT成果登Nature正刊：90天，「AI科学家」完成3500次电化学测试
MIT成果登Nature正刊：90天，「AI科学家」完成3500次电化学测试
AGILE：视觉学习新范式！自监督+交互式强化学习助力VLMs感知与推理全面提升
AGILE：视觉学习新范式！自监督+交互式强化学习助力VLMs感知与推理全面提升
Vision-Zero：零数据VLM自我进化！陈怡然团队提出零监督训练新范式
欧几里得的礼物：通过几何代理任务增强视觉-语言模型中的空间感知和推理能力
Method teaches generative AI models to locate personalized objects
Method teaches generative AI models to locate personalized objects
When Does Supervised Training Pay Off? The Hidden Economics of Object Detection in the Era of Vision-Language Models
Towards Self-Refinement of Vision-Language Models with Triangular Consistency
Vision-Zero：零数据VLM自我进化！陈怡然团队提出零监督训练新范式
东京大学、牛津大学等联合发布VLA万字综述：机器人迈向通用智能的全栈指南
Your Vision-Language Model Can't Even Count to 20: Exposing the Failures of VLMs in Compositional Counting
