GAN Dissection: Visualizing and Understanding Generative Adversarial Networks
Antonio Torralba
Delta Electronics Professor of Electrical Engineering and Computer Science; Head, Faculty of AI and Decision-making, MIT EECS, MIT Schwarzman College of Computing

Antonio Torralba is the Delta Electronics Professor of Electrical Engineering and Computer Science at MIT and an investigator at the Computer Science and Artificial Intelligence Laboratory. He also heads the faculty of artificial intelligence and decision-making in the MIT Schwarzman College of Computing. Previously, he led the MIT Quest for Intelligence as its inaugural director, and was the MIT director of the MIT–IBM Watson AI Lab. Torralba researches computer vision, machine learning, and human visual perception, with an interest in building systems that can perceive the world the way humans do. He has received an NSF Career award, the International Association for Pattern Recognition’s JK Aggarwal Prize, a Frank Quick Faculty Research Innovation Fellowship and a Louis D. Smullin (’39) Award for Teaching Excellence. Torralba earned a BS from Telecom BCN, Spain, and a PhD from the Institut National Polytechnique de Grenoble, France.
Selected Publications
- Gan, C., Xian, Z., Zhu, B., Xu, Z., Tung, H-Y., Torralba, A., Fragkiadaki, K. (2023). FluidLab: A Differentiable Environment for Benchmarking Complex Fluid Manipulation. International Conference on Learning Representations (ICLR).
- Su, K., Qian, K., Shlizerman, E., Torralba, A., Chuang Gan, C. (2023). Physics-Driven Diffusion Models for Impact Sound Synthesis from Videos. Conference on Computer Vision and Pattern Recognition (CVPR).
- Gan, C., Gutfreund, D., Schwartz, J., McDermott, J., Tenenbaum, J., Alter, S., Zhou, S., Torralba, A., Gu, Y., Traer, J. (2022). Finding Fallen Objects Via Asynchronous Audio-Visual Integration. Conference on Computer Vision and Pattern Recognition (CVPR).
- Baradad, M., Chen, C-F., Wulff, J., Wang, T., Feris, R., Torralba, A., Isola, P. (2022). Procedural Image Programs for Representation Learning. Conference on Neural Information Processing Systems (NeurIPS).
Media
- May 24, 2023: MIT News, Helping robots handle fluids.
- November 23, 2022: MIT News, A simpler path to better computer vision.
- November 8, 2019: MIT News, Visualizing an AI model’s blind spots.
- July 1, 2019: MIT News, Teaching artificial intelligence to create visuals with more common sense.
- January 10, 2019: MIT Technology Review, A neural network can learn to organize the world it sees into concepts—just like we do.
- August 30, 2018: Quanta magazine, The new science of seeing around corners.
Publications with the MIT-IBM Watson AI Lab