A
Antonio Torralba
Researcher at Massachusetts Institute of Technology
Publications - 437
Citations - 105763
Antonio Torralba is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Computer science & Object detection. The author has an hindex of 119, co-authored 388 publications receiving 84607 citations. Previous affiliations of Antonio Torralba include Vassar College & Nvidia.
Papers
More filters
Posted Content
Image GANs meet Differentiable Rendering for Inverse Graphics and Interpretable 3D Neural Rendering.
TL;DR: This paper aims to extract and disentangle 3D knowledge learned by generative models by utilizing differentiable renderers, and significantly outperforms state-of-the-art inverse graphics networks trained on existing datasets, both quantitatively and via user studies.
Posted Content
Learning Compositional Koopman Operators for Model-Based Control
TL;DR: This paper proposes to learn compositional Koopman operators, using graph neural networks to encode the state into object-centric embeddings and using a block-wise linear transition matrix to regularize the shared structure across objects.
Proceedings ArticleDOI
Diverse Image Generation via Self-Conditioned GANs
TL;DR: A class-conditional GAN model without using manually annotated class labels is trained, conditional on labels automatically derived from clustering in the discriminator’s feature space, which outperforms several competing methods when addressing mode collapse.
Proceedings ArticleDOI
Accidental pinhole and pinspeck cameras: Revealing the scene outside the picture
TL;DR: Two types of “accidental” images that can be formed in scenes are identified and studied, including an accidental pinhole camera image that can reveal structures outside a room, or the unseen shape of the light aperture into the room.
Anticipating Visual Representations from Unlabeled Video
TL;DR: In this article, a framework that capitalizes on temporal structure in unlabeled video to learn to anticipate human actions and objects is presented. But this task is challenging partly because it requires leveraging extensive knowledge of the world that is difficult to write down.