Saeed Amizadeh

Saeed Amizadeh is a researcher in Microsoft’s Applied Sciences Group (ASG).  His research interests are mainly focused on neuro-symbolic visual reasoning and visual question answering.  Before joining ASG, Saeed was a member of the ML.NET team at Microsoft developing neuro-symbolic deep-learning methods for constraint-satisfaction problems as well as time-series anomaly-detection solutions for ML.NET.  Before that, he was a researcher at Yahoo Labs in Sunnyvale, Calif.

Saeed received his Ph.D. in artificial intelligence from the University of Pittsburgh; his M.S. in artificial intelligence and robotics, and B.S. in computer science are both from the University of Tehran.

Outside of work, Saeed likes dancing and DJing Argentine tango.

• CorrGAN: Simultaneous Learning of Speech Enhancement and Perceptual Quality Loss Functions

  To appear in ICASSP 2025 2025

  By Vasily Zadorozhnyy, Saeed Amizadeh, Qiang Ye, Kazuhito Koishida
• Weakly-Supervised Audio Separation via Bi-modal Semantic Similarity

  Proceedings of the Twelfth International Conference on Learning Representations (ICLR) May, 2024 Vol. abs/2404.01740

  By Tanvir Mahmud, Saeed Amizadeh, Kazuhito Koishida, Diana Marculescu
• Neuro-Symbolic Visual Reasoning: Disentangling "Visual" from "Reasoning"

  Proceedings of the 37th International Conference on Machine Learning (ICML), Vienna, Austria July, 2020

  By Saeed Amizadeh, Hamid Palangi, Oleksandr Polozov, Yichen Huang, Kazuhito Koishida
• PDP: A General Neural Framework for Learning Constraint Satisfaction Solvers

  CoRR March, 2019 Vol. abs/1903.01969

  By Saeed Amizadeh, Sergiy Matusevych, Markus Weimer
• Learning To Solve Circuit-SAT: An Unsupervised Differentiable Approach

  The 7th International Conference on Learning Representations (ICLR) 2019

  By Saeed Amizadeh, Sergiy Matusevych, Markus Weimer