题目:OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and Understanding
会议:Conference on Neural Information Processing Systems 2024
论文:http://arxiv.org/abs/2406.19389
主页:https://lxtgh.github.io/project/omg_llava/
年份:2024
单位:武汉大学等
等等。。。先不读了吧 这个就分割 不是双输入变化检测的 下次有空再读