Talks and media

Talks and media

Invited talks, research talks and posters by Ali Vosoughi.

Invited and research talks

  • Multimodal AI Research: From Benchmarks to Product. Invited research talk, Dolby Laboratories, January 2025.
  • Unified Audiovisual Understanding. Microsoft Research, audio and acoustics team, August 2024.
  • Counterfactual Audio-Language Learning. Bosch Center for Artificial Intelligence, July 2023.
  • Leveraging Pre-Images to Discover Nonlinear Relationships in Multivariate Environments. EUSIPCO 2021, Dublin.

Posters and workshop presentations

  • I^2: Generating Instructional Illustrations via Text-Conditioned Diffusion. CVPR 2026 Workshop on Generative AI for Storytelling (AISTORY), June 2026.
  • Learning Audio Concepts from Counterfactual Natural Language, and AVVA: Audio-Video Vector Alignment on Unlabeled Data. SANE 2024 (Speech and Audio in the Northeast) at Google, Cambridge, MA, October 2024.
  • Side-channel analysis with machine learning. SRC TECHCON 2019, selected poster.

PromptReverb was an oral paper at ICASSP 2026.