jfzhang95 / pytorch-video-recognition

PyTorch implemented C3D, R3D, R2Plus1D models for video activity recognition.
MIT License
1.18k stars 250 forks source link

It's a question where the accuracy is too low accuracy. accuracy(0.02~0.03) #67

Closed Eunchan24 closed 2 years ago

Eunchan24 commented 2 years ago

The c3d model's accuracy is too low. I've tried all the other ways to solve the issue, but it's not solved. jfzhang95/pytorch-video-recognition#32 jfzhang95/pytorch-video-recognition#50 jfzhang95/pytorch-video-recognition#23 Please help me.