Yushi Li

Yushi Li

Auto-generated profile, based on publicly available information

Yushi Li

Department of Intelligent Science, Xian Jiaotong-Liverpool University, Suzhou, China

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 April 2021 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
YLYushi Li
Yushi Li
GBGeorge Baciu
George Baciu
Point clouds are the most general data representations of real and abstract objects, and have a wide variety of applications in many science and engineering fields. Point clouds also provide the most scalable multi-resolution composition for geometric structures. Although point cloud learning has shown remarkable results in shape estimation and semantic segmentation, the unsupervised generation of 3D object parts still pose significant challenges in the 3D shape understanding problem. We address this problem by proposing a novel Generative Adversarial Network (GAN), named HSGAN, or Hierarchical Self-Attention GAN, with remarkable properties for 3D shape generation. Our generative model takes a random code and hierarchically transforms it into a representation graph by incorporating both Graph Convolution Network (GCN) and self-attention. With embedding the global graph topology in shape generation, the proposed model takes advantage of the latent topological information to fully construct the geometry of 3D object shapes. Different from the existing generative pipelines, our deep learning architecture articulates three significant properties HSGAN effectively deploys the compact latent topology information as a graph representation in the generative learning process and generates realistic point clouds, HSGAN avoids multiple discriminator updates per generator update, and HSGAN preserves the most dominant geometric structures of 3D shapes in the same hierarchical sampling process. We demonstrate the performance of our new approach with both quantitative and qualitative evaluations. We further present a new adversarial loss to maintain the training stability and overcome the potential mode collapse of traditional GANs. Finally, we explore the use of HSGAN as a plug-and-play decoder in the auto-encoding architecture.

26 March 2021 in IEEE Transactions on Visualization and Computer Graphics
Verified
Journal Article
Closed access
YLYushi Li
Yushi Li
GBGeorge Baciu
George Baciu
Point clouds are fundamental in the representation of 3D objects. However, they can also be highly unstructured and irregular. This makes it difficult to directly extend 2D generative models to three-dimensional space. In this article, we cast the problem of point cloud generation as a topological representation learning problem. In order to capture the representative features of 3D shapes in the latent space, we propose a hierarchical mixture model that integrates self-attention with an inference tree structure for constructing a point cloud generator. Based on this, we design a novel Generative Adversarial Network (GAN) architecture that is capable of generating recognizable point clouds in an unsupervised manner. The proposed adversarial framework (SG-GAN) relies on self-attention mechanism and Graph Convolution Network (GCN) to hierarchically infer the latent topology of 3D shapes. Embedding and transferring the global topology information in a tree framework allows our model to capture and enhance the structural connectivity. Furthermore, the proposed architecture endows our model with partially generating 3D structures. Finally, we propose two gradient penalty methods to stabilize the training of SG-GAN and overcome the possible mode collapse of GAN networks. To demonstrate the performance of our model, we present both quantitative and qualitative evaluations and show that SG-GAN is more efficient in training and it exceeds the state-of-the-art in 3D point cloud generation.

17 June 2020 in IEEE Access
Verified
Journal Article
Open access
ZLZhe Liang
Zhe Liang
CXChao Xu
Chao Xu
JHJing Hu
Jing Hu
YLYushi Li
Yushi Li
ZMZhaopeng Meng
Zhaopeng Meng
Passive reconstruction methods such as traditional multi-view stereo are capable to accurate reconstruction results. However, the depth calculation of multi-view stereo encounters significant difficulties, especially when the corresponding points have some degree of inaccuracy or unreliability. In this paper, we make use of geometric and shading cues information in the multi-view stereo approach to construct a robust energy function, which is also effective to optimize the depth information provided by multi-view stereo. This work improves the accuracy of point cloud depth. we evaluated our algorithm in the famous DTU datasets, and established our own dinosaur datasets. All the data is collected by mobile devices under natural light condition, while the reconstruction results are completed and accurate.

29 November 2023 in Computers & Education: X Reality
Verified
Journal Article
Open access
YLYiming Luo
Yiming Luo
SSShuling Sun
Shuling Sun
YWYihong Wang
Yihong Wang
Di WuDi Wu
Di Wu
YOYuanbing Ouyang
Yuanbing Ouyang
YPYushan Pan
Yushan Pan

27 August 2024 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
YLYushi Li
Yushi Li
GBGeorge Baciu
George Baciu
RCRong Chen
Rong Chen
CLChenhui Li
Chenhui Li
Hao WangHao Wang
Hao Wang
Weiping DingWeiping Ding
Weiping Ding
Learning 3-D structures from incomplete point clouds with extreme sparsity and random distributions is a challenge since it is difficult to infer topological connectivity and structural details from fragmentary representations. Missing large portions of informative structures further aggravates this problem. To overcome this, a novel graph convolutional network (GCN) called dynamic and structure-aware NETwork (DSANet) is presented in this article. This framework is formulated based on a pyramidic auto-encoder (AE) architecture to address accurate structure reconstruction on the sparse and incomplete point clouds. A PointNet-like neural network is applied as the encoder to efficiently aggregate the global representations of coarse point clouds. On the decoder side, we design a dynamic graph learning module with a structure-aware attention (SAA) to take advantage of the topology relationships maintained in the dynamic latent graph. Relying on gradually unfolding the extracted representation into a sequence of graphs, DSANet is able to reconstruct complicated point clouds with rich and descriptive details. To associate analogous structure awareness with semantic estimation, we further propose a mechanism, called structure similarity assessment (SSA). This method allows our model to surmise semantic homogeneity in an unsupervised manner. Finally, we optimize the proposed model by minimizing a new distortion-aware objective end-to-end. Extensive qualitative and quantitative experiments demonstrate the impressive performance of our model in reconstructing unbroken 3-D shapes from deficient point clouds and preserving semantic relationships among different regional structures.

6 March 2025 in Multimedia Systems
Verified
Journal Article
Closed access
PDF file
HCHui Chen
Hui Chen
RCRong Chen
Rong Chen
YLYushi Li
Yushi Li
HLHaoran Li
Haoran Li
NLNannan Li
Nannan Li