F
Fereshteh Sadeghi
Researcher at University of Washington
Publications - 30
Citations - 1693
Fereshteh Sadeghi is an academic researcher from University of Washington. The author has contributed to research in topics: Visual servoing & Computer science. The author has an hindex of 14, co-authored 26 publications receiving 1400 citations. Previous affiliations of Fereshteh Sadeghi include Amirkabir University of Technology & University of Central Florida.
Papers
More filters
Proceedings ArticleDOI
CAD2RL: Real Single-Image Flight without a Single Real Image
Fereshteh Sadeghi,Sergey Levine +1 more
TL;DR: In this article, a collision avoidance policy is represented by a deep convolutional neural network that directly processes raw monocular images and outputs velocity commands, with a Monte Carlo policy evaluation algorithm that directly optimizes the network's ability to produce collision free flight.
Posted Content
CAD2RL: Real Single-Image Flight without a Single Real Image
Fereshteh Sadeghi,Sergey Levine +1 more
TL;DR: In this paper, a collision avoidance policy is represented by a deep convolutional neural network that directly processes raw monocular images and outputs velocity commands, with a Monte Carlo policy evaluation algorithm that directly optimizes the network's ability to produce collision free flight.
Proceedings ArticleDOI
Sim2Real Viewpoint Invariant Visual Servoing by Recurrent Control
TL;DR: In this article, a deep recurrent controller is trained to automatically determine which actions move the end-effector of a robotic arm to a desired object by using its memory of past movements, correcting mistakes and gradually moving closer to the target.
Proceedings ArticleDOI
VisKE: Visual knowledge extraction and question answering by visual verification of relation phrases
TL;DR: This work introduces the problem of visual verification of relation phrases and developed a Visual Knowledge Extraction system called VisKE, which has been used to not only enrich existing textual knowledge bases by improving their recall, but also augment open-domain question-answer reasoning.
Book ChapterDOI
Latent pyramidal regions for recognizing scenes
TL;DR: The proposed LPR representation obtains state-of-the-art results on all these datasets which shows that it can simultaneously model the global and local scene characteristics in a single framework and is general enough to be used for both indoor and outdoor scene classification.