Assignment: https://www.fit.vutbr.cz/study/courses/SUR/public/projekt_2019-2020/SUR_projekt2019-2020.txt
The program attempts to match an audio sample of a person's voice and an image of their face to the corresponding person.
For audio detection we chose the HMM clasificator and for image recogniction we chose the lightgbm clasificator.
Required packages and their version can be found in the requirements.txt file.
If any packages are missing or are of an incorrect version please run:
pip install -r requirements.txt
If .model files are missing or the models need re-training
python make.py train
For training place training data in the target_train/ or non_target_train/ folders based on whether the data is to be expected to be a hit or not.
To run evaluation:
python make.py eval
For evaluation the required data is to be placed in eval/ folder