Jinrong Zhang

Results 5 issues of Jinrong Zhang

Hello, it seems that there is a problem with the Huggingface Demo you provided and it cannot run properly. https://huggingface.co/spaces/ShilongLiu/Grounding_DINO_demo ![image](https://github.com/IDEA-Research/GroundingDINO/assets/49579978/d2c4a010-d6d4-4451-9532-1019ce32c4e3)

Could you give a table corresponding to joints and labels of the human3.6m and CMU dataset? 请问可以给一下h3.6m和CMU数据集的关节编号对应关机位置的表格吗

您好,您这真的是一篇很棒的工作!我有一些自己的实例分割和目标检测的数据集,想要在您的工作上面尝试微调。能否给出一些在您的工程上配置自己训练数据的更详细的步骤?非常感谢!

请问改如何将您提供的参数修改为能在单卡上进行推理的版本? 我现在的测试脚本是: python tools/train_net.py \ --eval-only \ --num-gpus 1 \ --resume \ --config-file configs/ape_deta/ape_train.py \ train.output_dir=output/val/odinw_MountainDewCommercial_test \ train.init_checkpoint=output/coco_06_24_v2/model_best.pth 报错显示: AttributeError: 'SomeThing' object has no attribute 'module'

您好,想问一下,ODinW35数据集中35个小数据集的结果是怎么算最后的总的结果的?是35个小数据集分别求了AP以后一起算35个数的平均值,还是把所有图片和所有实例目标何在一起算AP的?