H
Haisheng Su
Researcher at Shanghai Jiao Tong University
Publications - 20
Citations - 953
Haisheng Su is an academic researcher from Shanghai Jiao Tong University. The author has contributed to research in topics: Context (language use) & Boundary (topology). The author has an hindex of 8, co-authored 18 publications receiving 463 citations. Previous affiliations of Haisheng Su include Huazhong University of Science and Technology & SenseTime.
Papers
More filters
Book ChapterDOI
BSN: Boundary Sensitive Network for Temporal Action Proposal Generation
TL;DR: An effective proposal generation method, named Boundary-Sensitive Network (BSN), which adopts "local to global" fashion and significantly improves the state-of-the-art temporal action detection performance.
Proceedings ArticleDOI
Temporal Context Aggregation Network for Temporal Action Proposal Refinement
Zhiwu Qing,Haisheng Su,Weihao Gan,Dongliang Wang,Wei Wu,Xiang Wang,Yu Qiao,Junjie Yan,Changxin Gao,Nong Sang +9 more
TL;DR: A Local-Global Temporal Encoder (LGTE) and Temporal Boundary Regressor (TBR) are designed to combine these two regression granularities in an end-to-end fashion, which achieves the precise boundaries and reliable confidence of proposals through progressive refinement.
Posted Content
BSN++: Complementary Boundary Regressor with Scale-Balanced Relation Modeling for Temporal Action Proposal Generation
TL;DR: BSN++ is presented, a new framework which exploits complementary boundary regressor and relation modeling for temporal proposal generation and introduces the scale-balanced re-sampling strategy.
Posted Content
BSN: Boundary Sensitive Network for Temporal Action Proposal Generation
TL;DR: Wang et al. as mentioned in this paper proposed a boundary-sensitive network (BSN) to generate temporal action proposals with high recall and high overlap using relatively fewer proposals, which adopts a "local to global" fashion.
Journal ArticleDOI
Attention-Based Multiview Re-Observation Fusion Network for Skeletal Action Recognition
TL;DR: An attention-based multiview re-observation fusion model for skeletal action recognition focusing on the factor of observation view of actions, which greatly influences action recognition, and stacking multiple layers of attention operation in a multilayer LSTM network to further improve network performance.