As part of the special sessions for the upcoming AES International Conference on Artificial Intelligence and Machine Learning for Audio (AIMLA 2025), we are excited to announce the AI Challenge on Sound Retrieval!
Challenge: Querying Sound Effects by Vocal Imitation
Website: https://qvim-aes.github.io/
Key Dates:
Start: April 1, 2025
End: June 15, 2025
Results Announcement: July 15, 2025
Prizes:
1st Place: $1200 USD
2nd Place: $700 USD
3rd Place: $300 USD
Challenge Description
This challenge invites participants to develop a system that enables Query by Vocal Imitation—allowing users to search a sound database by recording a vocal impression of the desired sound. The system should retrieve perceptually similar recordings, offering sound designers an intuitive way to navigate large sound effects databases.
Evaluation: A subjective evaluation using a larger, unlabeled dataset will determine the final rankings of submissions.
Interested in Participating?
Visit the challenge website for details and submission guidelines:
https://qvim-aes.github.io/
We look forward to your participation!