研究方向为计算机视觉和多模态人工智能，专注于词元压缩，即在保持模型性能的同时减少多模态模型中的视觉词元数量。
He is working on computer vision and multimodal AI, focusing on token compression to reduce the number of visual tokens in multimodal models while maintaining model performance.
研究方向为计算机视觉和多模态人工智能，专注于词元压缩，即在保持模型性能的同时减少多模态模型中的视觉词元数量。
He is working on computer vision and multimodal AI, focusing on token compression to reduce the number of visual tokens in multimodal models while maintaining model performance.
