Y
Yangqing Jia
Researcher at Facebook
Publications - 61
Citations - 93683
Yangqing Jia is an academic researcher from Facebook. The author has contributed to research in topics: Deep learning & Image segmentation. The author has an hindex of 37, co-authored 61 publications receiving 78214 citations. Previous affiliations of Yangqing Jia include Tsinghua University & Google.
Papers
More filters
Proceedings ArticleDOI
Beyond spatial pyramids: Receptive field learning for pooled image features
TL;DR: This paper shows that learning more adaptive receptive fields increases performance even with a significantly smaller codebook size at the coding layer, and adopts the idea of over-completeness to learn the optimal pooling parameters.
Journal ArticleDOI
Trace Ratio Problem Revisited
TL;DR: A theoretical overview of the global optimum solution to the TR problem via the equivalent trace difference problem is proposed, and Eigenvalue perturbation theory is introduced to derive an efficient algorithm based on the Newton-Raphson method.
Posted Content
Deep Convolutional Ranking for Multilabel Image Annotation
TL;DR: In this paper, a significant performance gain could be obtained by combining convolutional architectures with approximate top-k$ ranking objectives, as they naturally fit the multilabel tagging problem.
A Category-Level 3D Object Dataset: Putting the Kinect to Work.
Allison Janoch,Sergey Karayev,Yangqing Jia,Jonathan T. Barron,Mario Fritz,Kate Saenko,Trevor Darrell +6 more
TL;DR: A dataset of color and depth image pairs, gathered in real domestic and office environments, establishes baseline performance in a PASCAL VOC-style detection task, and suggests two ways that inferred world size of the object may be used to improve detection.
Proceedings Article
Factorized Latent Spaces with Structured Sparsity
TL;DR: This paper shows that structured sparsity allows us to address the multi-view learning problem by alternately solving two convex optimization problems and shows that the resulting factorized latent spaces generalize over existing approaches in that they allow having latent dimensions shared between any subset of the views instead of between all the views only.