https://github.com/willprice/two-stream-action-cnn-analysis
:mortar_board: Undergraduate thesis on analysing two stream CNNs for action recognition using excitation backpropagation
https://github.com/willprice/two-stream-action-cnn-analysis
action-recogntion attribution caffe cnn convolutional-neural-networks dnn ebp python visualisation visualization
Last synced: about 2 months ago
JSON representation
:mortar_board: Undergraduate thesis on analysing two stream CNNs for action recognition using excitation backpropagation
- Host: GitHub
- URL: https://github.com/willprice/two-stream-action-cnn-analysis
- Owner: willprice
- Created: 2016-11-25T11:39:40.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2018-03-27T13:36:31.000Z (about 8 years ago)
- Last Synced: 2025-10-12T09:06:39.017Z (8 months ago)
- Topics: action-recogntion, attribution, caffe, cnn, convolutional-neural-networks, dnn, ebp, python, visualisation, visualization
- Language: Jupyter Notebook
- Homepage:
- Size: 39.3 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 2SCNN Feature analysis
Working repository for feature analysis of 2SCNN
## Models
* [CUHK VGG16-2SCNN models trained on UCF101
](https://github.com/yjxiong/caffe/tree/action_recog/models/action_recognition)
* UoB VGG16-2SCNN trained on BEOID (not publicly available)
## Dependencies
* Python 3
* matplotlib
* numpy
* scipy
* pandas
* seaborn
* click
* skimage
* Jupyter notebook (6.0)
* IPykernel for Jupyter (5.0)
* Bash
* [Caffe (Excitation BP fork)](https://github.com/jimmie33/Caffe-ExcitationBP),
preferably built with GPU support.
## Setup
* Edit `lib/config.py` to point to your local [excitation bp
caffe](https://github.com/jimmie33/Caffe-ExcitationBP) installation,
nets path (e.g. `caffe/models`, I like to keep mine in my home directory), and
data root.
## Repository overview
* `lib`: Most of the meaty code lives in here, code for handling datasets,
performing contrastive EBP, mapping from class ids to class names.
* `net_configs`: Python library of configuration scripts defining settings
peculiar to models.
* `scripts`: Python CLI scripts for generating attention maps, stitching videos,
graphing smoothness
* `generated`: A few bash scripts live in here to call the python scripts that
do most of the work.
* `notebooks`: All the Jupyter notebooks used for experimental analysis reside
here. They're meant to be run in numerical order. Notebooks with the same
numerical prefix can be run in any order.
* `./run_jupyter.sh` is a helper script to setup my environment with the correct
`LD_LIBRARY_PATH` and `PATH` env variables necessary to find Caffe, CuDNN etc.
## Acknowledgments
The code for performing contrastive excitation backprop is from
https://github.com/jimmie33/Caffe-ExcitationBP, I also took inspiration from
the code used to overlay attention maps on images.