M
Mingyuan Zhang
Researcher at Beihang University
Publications - 21
Citations - 593
Mingyuan Zhang is an academic researcher from Beihang University. The author has contributed to research in topics: Computer science & Deep learning. The author has an hindex of 6, co-authored 14 publications receiving 147 citations. Previous affiliations of Mingyuan Zhang include SenseTime.
Papers
More filters
Posted Content
Efficient Attention: Attention with Linear Complexities.
TL;DR: A novel efficient attention mechanism is proposed, which is equivalent to dot-product attention but has substantially less memory and computational costs and democratizes attention to complicated models, which were unable to incorporate original dot- product attention due to prohibitively high costs.
Journal ArticleDOI
Spatio-temporal deformable 3D ConvNets with attention for action recognition
TL;DR: This paper proposes a spatio-temporal deformable ConvNet module with an attention mechanism, which takes into consideration the mutual correlations in both temporal and spatial domains, to effectively capture the long-range and long-distance dependencies in the video actions.
Journal ArticleDOI
Spatio-Temporal Attention Networks for Action Recognition and Detection
TL;DR: A spatio-temporal attention (STA) network that is able to learn the discriminative feature representation for actions, by respectively characterizing the beneficial information at both the frame level and the channel level to enhance the learning capability of the 3D convolutions when handling the complex videos.
Journal ArticleDOI
AvatarCLIP: Zero-Shot Text-Driven Generation and Animation of 3D Avatars
TL;DR: By leveraging the priors learned in the motion VAE, a CLIP-guided reference-based motion synthesis method is proposed for the animation of the generated 3D avatar, which validate the effectiveness and generalizability of texture generation.
Journal ArticleDOI
Graph Attention Based Proposal 3D ConvNets for Action Detection
TL;DR: This work proposes graph attention based proposal 3D ConvNets (AGCN-P-3DCNNs), a simple and effective framewise classifier, which enhances the feature presentation capabilities of backbone model and demonstrates the state-of-the-art performance achieved.