In this challenge, you will need to create solutions that can tell apart real and fake audio recordings of clients. You will develop tools to check if the information shared in these calls matches the client's profile. Additionally, you will need to figure out if a human voice in the recording is the actual client or an impersonator. We will provide you with a set of client profiles each with one verified authentic sample of the client's voice. Furthermore, you will receive the labels for a subset of the dataset and we expect you to provide the labels for the remaining audio clips (more details follow below).
- Identify deepfake client calls
- Verify the client's identity on the factual content of the call
- Differentiate between distinct human voices