J
Jia Deng
Researcher at Princeton University
Publications - 158
Citations - 110718
Jia Deng is an academic researcher from Princeton University. The author has contributed to research in topics: Computer science & Object detection. The author has an hindex of 50, co-authored 148 publications receiving 73461 citations. Previous affiliations of Jia Deng include University of Michigan & Carnegie Mellon University.
Papers
More filters
Posted Content
D3D: Distilled 3D Networks for Video Action Recognition
TL;DR: This work investigates whether motion representations are indeed missing in the spatial stream, and shows that there is significant room for improvement, and demonstrates that these motion representations can be improved using distillation, that is, by tuning the spatial streams to mimic the temporal stream, effectively combining both models into a single stream.
Proceedings Article
DeepV2D: Video to Depth with Differentiable Structure from Motion
Zachary Teed,Jia Deng +1 more
TL;DR: DeepV2D as discussed by the authors combines the representation ability of neural networks with the geometric principles governing image formation for predicting depth from video. But it is not an end-to-end architecture.
Posted Content
RAFT-3D: Scene Flow using Rigid-Motion Embeddings.
Zachary Teed,Jia Deng +1 more
TL;DR: RAFT-3D is introduced, a new deep architecture for scene flow based on the RAFT model developed for optical flow but iteratively updates a dense field of pixelwise SE3 motion instead of 2D motion, which represents a soft grouping of pixels into rigid objects.
Posted Content
Learning to Prove Theorems via Interacting with Proof Assistants
Kaiyu Yang,Jia Deng +1 more
TL;DR: In this paper, a deep learning-based model that generates tactics as programs in the form of abstract syntax trees (ASTs) is proposed to prove new theorems not previously provable by automated methods.
Proceedings ArticleDOI
Surface Normals in the Wild
TL;DR: This paper collected human annotated surface normals and used them to help train a neural network that directly predicts pixel-wise depth, and proposed two novel loss functions for training with surface normal annotations.