For any queries email: s4p.daiict@gmail.com
Participants will work with the DeepFox 2.0: A Multi-Lingual Audio Deepfake Corpus, one of the largest and most diverse audio datasets developed to date. This dataset includes over 1,125 hours of audio spanning 20 languages, with more than 400,000 audio files. The data includes both fully synthesized audio and novel “half-truth” samples, where segments of real audio are replaced with deepfakes generated using 51 state-of-the-art models, which includes generation of deepfakes through variety of methods, such as (GANs, LLMs, VCs, and TTS). These features make the task of detection particularly challenging and realistic.
Compete for a prize of ₹22,500 (1st -> 10K, 2nd -> 7.5K, and 3rd -> 5K) and the opportunity to contribute to cutting-edge research while gaining hands-on experience with a highly challenging dataset. Whether you're a researcher, developer, or enthusiast, this challenge offers a unique platform to demonstrate your expertise and make a significant impact in the field of speech and signal processing.
Please visit the Deepfake Challenge Website here!
