/DLCVFinalProject_TalkingToMe

Primary LanguagePython

DLCV Final Project ( Talking to me ) | Poster

├──student_data
    ├──test
    ├──train
    ├──videos
    ├──Frame
        ├──{video_hashcode}_{frame_id}.jpg
    ├──Audio
        ├──{video_hashcode}.wav

Environment

# install the dependency
pip install -r requirements.txt

Before reproducing

# Download the model by executing
bash download.sh

Data Pre-processing

# Download the model by executing
bash prerocess.sh <Path to videos folder> <Path to seg folder> <Path to bbox folder>
# i.e. bash preprocess.sh './student_data/videos' './student_data/train/seg' './student_data/train/bbox' && bash preprocess.sh './student_data/videos' './student_data/test/seg' './student_data/test/bbox'

Reproduce the inference

# to reproduce test file
bash inference.sh <Path to videos folder> <Path to seg folder> <Path to bbox folder>  <Path to output csv file>
# i.e. bash inference.sh './student_data/videos' './student_data/test/seg' './student_data/test/bbox' 'output.csv'

Training

# to run the training
bash train.sh <Path to videos folder> <Path to seg folder> <Path to bbox folder>
# i.e. bash train.sh './student_data/videos' './student_data/train/seg' './student_data/train/bbox'

Contributors

  • 薛博文, Dobby-the-elf
  • 李尚宴, ShangYenLee
  • 許雅晴, Shelley1214
  • 黃秉茂, maomao0819