Sound Source Localization study for AI Grand Challenge 2021 (sponsored by NC Soft Vision Lab)
1. Place the data-listed csv file in the path below.
Sound-Source-Localization/ └── dataset/ └── dataset.csv
2. Run main.sh for training
cd Sound-Source-Localization/ sh train.sh
3. Run test.sh for test
cd Sound-Source-Localization/ sh test.sh
The results of the experiment are summarized in the link below.
We made cross-modal transformer model like below.