K
Kaili Zhao
Researcher at Beijing University of Posts and Telecommunications
Publications - 21
Citations - 920
Kaili Zhao is an academic researcher from Beijing University of Posts and Telecommunications. The author has contributed to research in topics: Facial expression & Cluster analysis. The author has an hindex of 10, co-authored 18 publications receiving 704 citations. Previous affiliations of Kaili Zhao include Peking University.
Papers
More filters
Proceedings ArticleDOI
Deep Region and Multi-label Learning for Facial Action Unit Detection
TL;DR: This paper proposes Deep Region and Multi-label Learning (DRML), a unified deep network that simultaneously addresses these two problems of facial Action Unit detection and region learning, allowing the two seemingly irrelevant problems to interact more directly.
Proceedings ArticleDOI
Joint patch and multi-label learning for facial action unit detection
TL;DR: This work introduces joint-patch and multi-label learning (JPML) to address issues of group sparsity and results show that in four of five comparisons on three diverse datasets, CK+, GFT, and BP4D, JPML produced the highest average F1 scores.
Journal ArticleDOI
Text extraction from natural scene image: A survey
TL;DR: This paper offers the researchers a link to public image database for the algorithm assessment of text extraction from natural scene images and draws attention to studies on the first two steps in the extraction process, since OCR is a well-studied area where powerful algorithms already exist.
Journal ArticleDOI
Joint Patch and Multi-label Learning for Facial Action Unit and Holistic Expression Recognition
TL;DR: A joint patch and multi-label learning (JPML) framework that models the structured joint dependence behind features, AUs, and their interplay, and can be extended to recognize holistic expressions by learning common and specific patches, which afford a more compact representation than the standard expression recognition methods.
Proceedings ArticleDOI
Learning Facial Action Units from Web Images with Scalable Weakly Supervised Clustering
TL;DR: A weakly-supervised spectral algorithm is derived that learns an embedding space to couple image appearance and semantics and offers intuitive outlier/noise pruning instead of forcing one annotation to every image.