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
Proceedings ArticleDOI
Procedural Image Programs for Representation Learning
Manel Baradad,Chun-Fu,Richard Chen,Jonas Wulff,Tong-Tong Wang,Rogerio Feris,Antonio Torralba,Phillip Isola +7 more
TL;DR: In this article , a large dataset of twenty-one thousand programs, each one generating a diverse set of synthetic images, is used for both supervised and unsupervised representation learning, and reduces the gap between pre-training with real and procedurally generated images by 38%.
Proceedings ArticleDOI
Unsupervised Non-parametric Geospatial Modeling from Ground Imagery
TL;DR: This paper introduces the first unsupervised method for automated hierarchical modeling of high-level latent regions using densely sampled, geotagged ground imagery at a global scale and shows the effectiveness of this method for discovering regional distributions at vastly different scales, including the Boston area, the United States, and the World.
Circuits neuromorphiques pour l'estimation du mouvement
Antonio Torralba,Jeanny Hérault +1 more
TL;DR: A new family of velocity-tuned filters much simpler than classical Gabor filters used in some motion estimation algorithms are presented, implemented with analogue networks with asymetrical lateral connexions which allows a complete control of the filter parameters.
Journal Article
Guest Editorial: Big Data
Alyosha Efros,Antonio Torralba +1 more
TL;DR: In computer vision, computer vision has a split personality as mentioned in this paper, and it combines two problems that are utterly disparate in their aims and philosophy: vision as measurement and vision as understanding.
Proceedings Article
Energy-Based Models for Continual Learning
TL;DR: Energy-Based Models (EBMs) have been shown to be adaptable to a more general continual learning setting where the data distribution changes without the notion of explicitly delineated tasks as discussed by the authors.