hulianyuyy / CorrNet

Continuous Sign Language Recognition with Correlation Network (CVPR 2023)
85 stars 14 forks source link

Trainning problem #7

Closed stevenGX24 closed 10 months ago

stevenGX24 commented 1 year ago

I tried to train on CSL-Daily dataset, but it reported the error "list index out of range". It may be the problem in dataloader_video.py. I notice that you said, we need to disable the temporal resampling strategy. How that actually been done? Should I comment the whole collate_fn function?

hulianyuyy commented 1 year ago

No, this is mostly attributed to that you haven't correctly set the path to CSL-Daily dataset.

---Original--- From: @.> Date: Tue, Jun 6, 2023 14:30 PM To: @.>; Cc: @.***>; Subject: [hulianyuyy/CorrNet] Trainning problem (Issue #7)

I tried to train on CSL-Daily dataset, but it reported the error "list index out of range". It may be the problem in dataloader_video.py. I notice that you said, we need to disable the temporal resampling strategy. How that actually been done? Should I comment the whole collate_fn function?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.***>

stevenGX24 commented 1 year ago

Do you mean I need to set the soft link to the sentence file?

hulianyuyy commented 1 year ago

Set the soft link to the dataset dir. You can follow the instructions for CSL-Daily dataset in the readme.md

---Original--- From: @.> Date: Tue, Jun 6, 2023 14:52 PM To: @.>; Cc: @.**@.>; Subject: Re: [hulianyuyy/CorrNet] Trainning problem (Issue #7)

Do you mean I need to set the soft link to the sentence file?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

stevenGX24 commented 1 year ago

OK, thanks.