Audio Deepfake Detection Systems
Reproduction of state-of-the-art audio deepfake detection systems, including LCNN, RawNet2, and AASIST.
Project GitHub: Not Available for teaching purposes. Year: 2023
The project aimed on the reproduction of LightCNN (LCCN) following the Speech Technology Center paper, RawNet2, and AASIST audio deepfake detection systems. The AASIST solution was written using PyTorch Geometric and Deep Graph Library to supplement knowledge in the application of libraries for graph neural networks. The project was developed to ensure that HSE DLA Course Homework is doable by students in a reasonable time.