Skip to content

Latest commit

 

History

History
37 lines (27 loc) · 858 Bytes

README.md

File metadata and controls

37 lines (27 loc) · 858 Bytes

Guitar - Augmented Reality - Deep Learning

  • Guitar fretboard detection (using Deep Learning - UNet)
  • Projecting a mesh on the detected fretboard (using Homography)
  • Displaying the notes to play on the projected mesh

Environment Setup

# Docker container to run train.py and prediction.py
./docker_start_pred.sh

# Docker container to run main.py
./docker_start.sh

Dataset

# Add training images to data/guitar/train/image/
# Add corresponding training mask images to data/guitar/train/labels/

# Use data_aug_offline.py to augment dataset offline[Optional]
pyhon3 data_aug_offline.py

Details

  • Train the model using train.py
  • Predict fretboard segmentation using prediction.py
  • Project mesh on to the guitar fretboard using main.py

Results

References

https://github.com/zhixuhao/unet