Step4-v2 less than 1 minute read Published: March 17, 2025Step4-v2这里我是想用相同的label来表示限制合并的遍历范围，但是这里的label会不会因为 mask3d的分割问题导致不一致有误呢？ （目前姑且认为下游推理是强大的综合的力挽狂澜的上下文认知深刻的）后面再做消融实验来改变我在犹豫是直接挑出来0704_00做实验还是批量输出和gt相比 尺寸错误的频率高低？我觉得先看频率吧…先尝试一个小时行不行Share on X (formerly Twitter) Facebook LinkedIn Previous Next
UCSD_Offer / New chapter less than 1 minute read Published: February 25, 2025Offer?! Really我无疑是幸运至极的。
UCSD_Meeting / Oh_My_Captain less than 1 minute read Published: February 24, 2025? Meeting ? Chat with Dr. Sun这是第一篇，刚打几个字就开始不想遵守标点符号规则了。
