arXiv · ML· Qize Yu, Lianrui Fan, Bowen Ping, Xini Ding, Zetian Song, Junbo Niu, Kaixuan Wang, Tianxing Chen, Yue Chen, Minghua He, Yuran Wang, Jie Huang, Haojun Zhang, Min Chen, Hao Li, Wenxuan Song, Ruihai Wu, Xianming Liu, Shilong Liu, Shuchang Zhou, Ping Luo, Shiyu Huang·· 8 小时前AI 评分27
GroundAnything：以 Flash 速度实现精确视觉定位与并行解码的调和
GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed
AI 导读
GroundAnything 4B 参数模型通过分块去噪实现快速并行解码，GroundAnything-VLM 在 30 项基准测试中取得 72.42% 平均成绩，达到同规模模型 SOTA。
来源：arXiv · ML · arxiv.org