ALM: Audio Language and Multimodal

ALM is a collaborative research group focused on deep learning for audio, language, and multimodal data.

About Us

• Alkis Koudounas - PhD Student at Politecnico di Torino (Profile | polito.it)
• Lorenzo Vaiani - PhD Student at Politecnico di Torino (Profile | polito.it)
• Moreno La Quatra - Research Fellow at Kore University of Enna (Profile | unikore.it)

Projects

• ARCH - Audio Representation Benchmark (Repo): A platform dedicated to benchmarking models for audio representations. Resaerch Paper
• CALM - Contrastive Alignment of Language and Music: A project from the 1st Sound of AI Hackathon. CALM aligns songs with natural language descriptions, enabling music searches via text, voice, or facial expressions.
• PACE - Podcast AI for Chapters and Episodes: PACE is a semantic search engine for podcasts. It enables users to search for specific parts of a podcast using natural language. The project was created for the AssemblyAI 50K Hackathon - Winter 2022.