search

xlliu7 / E2E-TAD

[CVPR 2022] An Empirical Study of End-to-end Temporal Action Detection
Apache License 2.0
81 stars 11 forks source link

Training By Myself #6

Open Hank5120 opened 1 year ago

Hank5120 commented 1 year ago

Hello, I want to train model by myself, But I don't know which dataset you use and the train code you use. Will you release this part? thanks

Hank5120 commented 1 year ago

I add another image imput, but it will error because the channel size is different, how can I do to adjust this part? Screenshot from 2022-12-14 17-23-46