H
Houqiang Li
Researcher at University of Science and Technology of China
Publications - 612
Citations - 17591
Houqiang Li is an academic researcher from University of Science and Technology of China. The author has contributed to research in topics: Computer science & Motion compensation. The author has an hindex of 57, co-authored 520 publications receiving 12325 citations. Previous affiliations of Houqiang Li include China University of Science and Technology & Nanjing Medical University.
Papers
More filters
Posted Content
Instance-wise Hard Negative Example Generation for Contrastive Learning in Unpaired Image-to-Image Translation
TL;DR: Zhang et al. as discussed by the authors proposed an instance-wise hard negative example generation for contrastive learning in unpaired image-to-image translation. But the negative examples are randomly sampled from the patches of different positions in the source image, which is not effective to push the positive examples close to the query examples.
Proceedings ArticleDOI
Parallel Encoding - Decoding Operation for Multiview Video Coding with High Coding Efficiency
TL;DR: This paper presents a novel coding structure that enables parallel encoder/decoder operation for different views, without compromising from the coding efficiency, and achieves up to 0.9 dB gain compared to simulcast.
Posted Content
Relation Distillation Networks for Video Object Detection
TL;DR: Relation Distillation Networks is presented --- a new architecture that novelly aggregates and propagates object relation to augment object features for detection and achieves superior results when comparing to state-of-the-art methods.
Proceedings ArticleDOI
Deep network-based image coding for simultaneous compression and retrieval
TL;DR: A deep network-based image coding scheme that achieves a compression ratio of 5.3 for 32×32 thumbnails, outperforms JPEG at similar compression ratios, and the resulting code is directly available for CBIR is indicated.
Journal Article
Face recognition using neighborhood preserving projections
TL;DR: In this paper, a novel unsupervised subspace learning method, Neighborhood Preserving Projections (NPP), is proposed, which has good neighborhood-preserving property and can be used for face recognition.