cv

Installation

To install this extension, first you need to clone this repo

pip install -r requirements.txt

To download detector and classifier checkpoint, click here

Usage

1. You need to run this command to open API:

python api.py --detector_path path/to/detector/checkpoint --classifier_path path/to/classifier/checkpoint

2. Open chrome browser extension and open developer mode. Click Load unpacked and select the directory you cloned
   
3. Open a random youtube video and click the extension icon -> Movie Controller
   
4. Tick to checkbox and enjoy your video
   

Training resnet18, resnext101, mobilenetv3_smal, vit32

1. To start training, run:

   python -m classifier.run --command 'train' --path_to_config classifier/config/*.yaml
   

Training MAE VIT

1. If you want to run ViT without MAE pretrain, just run

   python VIT_classifier.py
   

2. If you want to run ViT with MAE pretrain, First, run:

   python MAE.py
   

   After finishing training, load the MAE checkpoint in the VIT_classifier.py file, and run:

   python VIT_classifier.py