Query for Pre-training Details

[bruceyo]

Hi Kyle Min, thank you for releasing the code and pre-trained model.

Could you kindly provide the training details like below for reproducing the accuracy on Kinetics? Thank you.

1. image size, is it resized to 224384 (heightwidth);
2. number of video frames, is it 32 frames, or 64 frames, or whole video frames;
3. do you use normalization by using the mean ([0.485, 0.456, 0.406]) and std ([0.229, 0.224, 0.225]).

Right now, without normalization and follow your code (i.e., the pre-processing in the main.py), I could get around 55% of accuracy on Kinetics 400 by using the image size of 224*384 and 32 frames.

Thank you in advance and appreciate your excellent work.

[kylemin]

Hi,
Thank you for your interest.

We followed the standard convention of Deepmind/I3D. Please refer to the previous discussion: link. All the training/testing details are in the original I3D paper.

[bruceyo]

Hi,
Thank you for your interest.

We followed the standard convention of Deepmind/I3D. Please refer to the previous discussion: link. All the training/testing details are in the original I3D paper.

Thank you for the kind reply. I note that the I3D paper uses the whole video. I also found that using more video frames leads to better performance. Thank you again :)