Jia-Wei Yan
National Taiwan University
5 Papers
2 Citations
Jia-Wei Yan is an academic researcher from National Taiwan University. The author has contributed to research in topics: Semantics & MNIST database. The author has an hindex of 1, co-authored 3 publications.
Chat about Author
Papers
Improving Subgraph Representation Learning via Multi-View Augmentation
TL;DR: In this article , a multi-view augmentation mechanism was proposed to improve the accuracy of downstream prediction tasks in subgraph representation learning, which creates multiple variants of subgraphs and embeds these variants into the original graph to achieve highly improved training efficiency and scalability.
2
•Posted Content
Semantics-Guided Representation Learning with Applications to Visual Synthesis
TL;DR: This paper proposes an angular triplet-neighbor loss (ATNL) that enables learning a latent representation whose distribution matches the semantic information of interest and utilizes spherical semantic interpolation for generating semantic warping of images, allowing synthesis of desirable visual data.
1
Semantics-Guided Representation Learning with Applications to Visual Synthesis
Jia-Wei Yan,Ci-Siang Lin,Fu-En Yang,Yu-Jhe Li,Yu-Chiang Frank Wang +4 more
- 10 Jan 2021
TL;DR: In this article, an angular triplet-neighbor loss (ATNL) is proposed to learn a latent representation whose distribution matches the semantic information of interest, which can then be used to generate semantic warping of images.
Deep Aggregation Net for Land Cover Classification
Tzu-Sheng Kuo,Keng-Sen Tseng,Jia-Wei Yan,Yen-Cheng Liu,Yu-Chiang Frank Wang +4 more
- 01 Jun 2018
TL;DR: A deep aggregation network is proposed for solving land cover classification, which extracts and combines multi-layer features during the segmentation process and introduces soft semantic labels and graph-based fine tuning in this proposed network for improving the segmentations performance.
Self-Supervised Category-Level 6D Object Pose Estimation with Deep Implicit Shape Representation
Wanli Peng,Jia-Wei Yan,Hongtao Wen,Yi Sun +3 more
TL;DR: A self-supervised framework for category-level 6D pose estimation that leverages DeepSDF as a 3D object representation and design several novel loss functions based onDeepSDF to help the self- supervised model predict unseen object poses without any 6D object pose labels and explicit 3D models in real scenarios.