This is the source code for our paper "Enhancing Single-Frame Supervision for Better Temporal Action Localization."
- The easiest way to install a environment to run the demo is to use docker. The image
chencjgene/actlocalizer-run:latest
contains the source codes and data, and have the dependencies installed. You can pull and run the image by:
$ docker pull chencjgene/actlocalizer-run:latest
- Run the docker image:
$ docker run -p 20222:20222 -p 30221:30221 -it chencjgene/actlocalizer-run:latest
- Run backend
$ cd /root/ActLocalizer/
$ nohup python manager.py run 30221 &
- Run frontend
$ cd vis
$ npm install (it will take a while)
$ npm run serve
- Visit http://localhost:20222/ in a browser.
-
This project uses python 3.8. Go check it out if you don't have it installed.
-
install python package.
$ pip install -r requirements.txt
$ pip install torch
-
Download the repo
-
Download demo data from here, and unpack it in the root folder
-
Run backend
$ nohup python manager.py run 30221 &
- Run frontend
$ cd vis
$ npm install (it will take a while)
$ npm run serve
- Visit http://localhost:20222/ in a browser.
If you have any problem about this code, feel free to contact
or describe your problem in Issues.