- To remove the monocolor background from instructor videos.
- To develop a robust algorithm that works in real-time
- The algorithm will work even when the background consists of real world scenes (curtains, windows, doors, etc)
Run the following commands to download the DeepLab models and install the necessary dependencies.
pip install requirements.txt
wget http://download.tensorflow.org/models/deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz
wget http://download.tensorflow.org/models/deeplabv3_pascal_train_aug_2018_01_04.tar.gz
mkdir mobile_net_model && tar xvzf deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz -C mobile_net_model
mkdir xception_model && tar xvzf deeplabv3_pascal_train_aug_2018_01_04.tar.gz
Segmentation can be performed on videos by running one of the following commands:
# Fast 25 FPS model
python seg_contour.py input_file.mp4 model 3 --video
# Accurate Model
python seg_contour.py input_file.mp4 model 2 --video
# Hybrid Heuristics based approach
python seg_v2.py ../green.mp4 --output_fpath out.avi --model_type x
This project was built in Bootstrap Paradox Hackathon organized by Blume Ventures. Also, our approach was considered in the top-5 list of all the ideas. Please click on this link to know more about the event.