Swathikiran Sudhakaran

No, I don't think I will try this model on EPIC-KITCHENS dataset.

Hi, EPIC-Kitchens dataset requires a different dataloader and multi-task training due to the structure of the labels and I didn't want to make the repo complicated. The implementation details are...

Hi, Please comment the script you used for training and the line where the error occurs. Also, is it occurring during validation or during training?

Hi, It seems like the error is triggered in somewhere that is not part of the original implementation shared in this repo. I am unable to help in this case....

`python3 main.py diving48 RGB --split 1 --arch InceptionV3 --num_segments 16 --consensus_type avg --batch-size 8 --iter_size 1 --dropout 0.7 --lr 0.01 --warmup 10 --epochs 20 --eval-freq 5 --gd 20 --run_iter 1...

Can you do a validation on the model shared to see if the same score reported in the paper is obtained. If not, I would suggest to extract the frames...

Can you extract the frames using the script given and try again?

I would assume so. I also had the same problem of getting very low scores when the frames were compressed.

I got between 36 and 39 in 4 or 5 runs. Class accuracy is the mean class accuracy. Overall prec@1 is the accuracy which is reported in sota methods.

But are you getting 40% accuracy with the shared model using 2 clips?