moabitcoin/ig65m-pytorch
PyTorch 3D video classification models pre-trained on 65 million Instagram videos
PythonMIT
Issues
- 5
r2plus1d_34_8_ig65m with 16 frames input
#39 opened by jaypatravali - 3
other pretrained models?
#44 opened by dragen1860 - 14
The models 'r2plus1d_34_clip32_ft_kinetics_from_ig65m' fine-tuned on UCF101 can't get good performance, just 78%, why?
#28 opened by fwcheng - 4
fine tune the pre-trained model on UCF101
#38 opened by Yueeeeee-1 - 1
- 2
can I download IG65M dataset?
#42 opened by pritamqu - 3
Citation information
#40 opened by Farabi-shafkat - 1
about extract features
#41 opened by kimsimple - 1
- 2
Provide CSN models and weights
#9 opened by daniel-j-h - 10
- 2
- 4
- 1
Need help with prediction codes
#34 opened by wuvei - 7
- 4
- 2
Script to convert caffe to pytorch
#32 opened by sanchit2843 - 0
Let CI build docker images
#14 opened by daniel-j-h - 0
Look into PyTorch Hub integration
#4 opened by daniel-j-h - 1
- 0
- 3
- 3
- 1
Add Travis CI integration
#13 opened by daniel-j-h - 0
Provide nvidia-docker based Dockerfile
#1 opened by daniel-j-h - 3
Port missing r(2+1)d 34-layer IG65-M model
#3 opened by daniel-j-h - 2