ImIntheMiddle
Not save the final model in the Activelearning.py
Hello, I noticed that the final model is not saved in your code. Initially, I thought lastest_AE.pth was the final model, but it turns out it isn't—this model is only used for sample screening. I wonder if you have noticed this issue.
Thanks for pointing this out! In Activelearning.py, since the fine-tuning and evaluation of a pose estimator is completed within the ATL process itself, we didn’t need to save the final model.
Unlike the super-lightweight AE, saving a pose estimator for every video across the entire dataset would take up a lot of storage. Of course, depending on your problem setting, you can save the pose estimator in the same way as the AE.
hi,I recently used the model of fastpose which had used dataset of posetrack to pretrain to fine-tune by the dataset of jrdb.The result of all method ,such as THC+WPU+DUW, the APs is near zero in the Round 0 .
