Скачать или смотреть Building a Hum-to-Search Music Recognition App with Vector Search

Join Sri Bala as she shares his experience of developing a unique music recognition app that can identify songs through humming or singing. In this detailed talk, she explores the challenges and solutions involved in building a hum-to-search application. Sri introduces various technologies used in the project including audio embeddings, specifically audio vector embeddings, and their applications in music recognition. She discusses the use of embedding models like PANS Inference and OpenL3 for detecting audio similarities.

The talk delves into MIDI data utilization for creating audio embeddings, emphasizing its compact and musical information-rich nature. Sri also demonstrates a live example of the app, illustrating its functionality in matching hummed tunes to actual songs through vector search within a database powered by AstroDB.

This video is an excellent resource for developers interested in AI, music technology, and innovative app development. Whether you’re an AI enthusiast looking to understand complex technical processes or simply curious about how music recognition technology works behind the scenes, this presentation provides valuable insights into the practical applications of AI in everyday technology.

#MusicTech #AI #AppDevelopment