fingerrec / 3dnet_visualization

Pytorch 3DNet attention feature map Visualization by [Cam](https://arxiv.org/abs/1512.04150); C3D, R3D, I3D, MF Net is support now!

3D Net Visualization Tools (PyTorch)

Demo

This project is to show which space-time region that the model focus on, supported supervised or unsupervised (no label available). For an input video, this project will show attention map in video and frames.

saved video

Video can't be show here, there are some gif.

supervised with label

unsupervised (only have RGB video)

saved img

heatmap

focus map

feature map average(without label)

In some case, the real label of video/action can't access. We average all filters and visualize the heatmap.

Require:

• pytorch0.4
• opencv
• numpy
• skvideo
• ffmpeg

Run:

1.create pretrain_model dir

git clone https://github.com/FingerRec/3DNet_Visualization.git
cd 3DNet_Visualization
mkdir pretrained_model

2.download pretrained model and put in into the dir pretrained_model

MF-Net

download pretrained MFNet on UCF101 from google_drive and put it into directory pretrained_model, which is from MFNet

I3d

google_drive

R3D

r3d

R3D pretrain model is from 3D-Resnet-Pytorch

C3D

C3D

C3D pretrain model is from C3D-Pytorch

3.run demo

pretrained I3d on HMDB51

bash scripts/demo.sh

c3d

bash scripts/c3d_unsupervised_demo.sh

r3d

bash scripts/r3d_unsupervised_demo.sh

The generate video and imgs will be put in dir output/imgs and output/video.

Tip: in main.py, if set clip_steps is 1, will generate a video the same length as origin.

4.test own video

the details in demo.sh as follow, change --video and --label accorading to your video, please refer to resources/classInd.txt for label information for UCF101 videos.

python main.py --num_classes 101 \
--classes_list resources/classInd.txt \
--model_weights pretrained_model/MFNet3D_UCF-101_Split-1_96.3.pth \
--video test_videos/[your own video here] \
--frames_num 16 --label 0 --clip_steps 16 \
--output_dir output \
--supervised unsupervised # not annotate this line if no label available

Notice unsupervised compute only add --supervised unsupervised in script;

Tip:UCF101/HMDB51 dataset is support now, for Kinetics et al. Just download a pretrained model and change --classes_list

To Do List

• support i3d, mpi3d
• support multi fc layers or full convolution networks
• support feature map average without label
• support r3d and c3d
• support Slow-Fast Net
• visualize filters
• grad-cam

More information

Support your own network:

1. pretrained model; 2. update load_model() in main.py; 3. modify last linear layer name in generate_supervised_cam in action_recognition.py

Notice C3D and R3D are pretrained on Sports/Kinetics, for better visualization, you may need to finetune these networks on UCF/HMDB as in RHE

Acknowledgment

This project is highly based on SaliencyTubes , MF-Net and st-gcn.

3dnet_visualization's Issues

3D ResNet and pre-trained model

good job！
Can you provide a code that includes 3D ResNet and a pre-trained model, directly using the two for visualization?

prob

How does the prob in the picture show up?

C3D visualization

hi，I saw in the to do list that you intend to support C3D visualization in this work. Have you do it right now？