Z
Zhuang Liu
Researcher at University of California, Berkeley
Publications - 53
Citations - 39804
Zhuang Liu is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Computer science & Artificial neural network. The author has an hindex of 25, co-authored 42 publications receiving 23096 citations. Previous affiliations of Zhuang Liu include Tsinghua University & Intel.
Papers
More filters
Proceedings ArticleDOI
Multi-Modal Contrastive Pre-training for Recommendation
TL;DR: A self-supervised contrastive inter-modal alignment task to make the textual and visual modalities as similar as possible as well as possible in order to exploit the potential correlation between users and items.
Proceedings ArticleDOI
SignInstructor: An Effective Tool for Sign Language Vocabulary Learning
TL;DR: This paper presents a natural and facilitate substitute, Sign Instructor, to help self-learning of SL vocabularies with an inexhaustible tutor, and convincingly shows the effectiveness in SL learning, especially compared with other two traditional ways.
Patent
Coarse-to-fine hand detection method using deep neural network
TL;DR: In this article, a detection process to identify one or more areas containing a hand or hands of one or multiple subjects in an image is provided, where the detection process can start with coarsely locating segments in the image that contain portions of the hand (s) of the subject (s).
Posted Content
Convolutional Networks with Dense Connectivity
TL;DR: The Dense Convolutional Network (DenseNet), which connects each layer to every other layer in a feed-forward fashion, and has several compelling advantages: they alleviate the vanishing-gradient problem, strengthen feature propagation, encourage feature reuse, and substantially improve parameter efficiency.
Journal ArticleDOI
MSeg: A Composite Dataset for Multi-Domain Semantic Segmentation
TL;DR: This work presents MSeg, a composite dataset that unifies semantic segmentation datasets from different domains, and adopts zero-shot cross-dataset transfer as a benchmark to systematically evaluate a model's robustness.