Hao Zhang

Hao Zhang

Auto-generated profile, based on publicly available information

Hao Zhang

School of Computing and Artificial Intelligence, Southwest Jiaotong University, Chengdu, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
20 September 2021 in IEEE Transactions on Cybernetics
Verified
Journal Article
Closed access
HZHao Zhang
Hao Zhang
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
TJTai-Xiang Jiang
Tai-Xiang Jiang
MNMichael NG
Michael NG
THTing-Zhu Huang
Ting-Zhu Huang
The general tensor-based methods can recover missing values of multidimensional images by exploiting the low-rankness on the pixel level. However, especially when considerable pixels of an image are missing, the low-rankness is not reliable on the pixel level, resulting in some details losing in their results, which hinders the performance of subsequent image applications (e.g., image recognition and segmentation). In this article, we suggest a novel multiscale feature (MSF) tensorization by exploiting the MSFs of multidimensional images, which not only helps to recover the missing values on a higher level, that is, the feature level but also benefits subsequent image applications. By exploiting the low-rankness of the resulting MSF tensor constructed by the new tensorization, we propose the convex and nonconvex MSF tensor train rank minimization (MSF-TT) to conjointly recover the MSF tensor and the corresponding original tensor in a unified framework. We develop the alternating directional method of multipliers (ADMMs) to solve the convex MSF-TT and the proximal alternating minimization (PAM) to solve the nonconvex MSF-TT. Moreover, we establish the theoretical guarantee of convergence for the PAM algorithm. Numerical examples of real-world multidimensional images show that the proposed MSF-TT outperforms other compared approaches in image recovery and the recovered MSF tensor can benefit the subsequent image recognition.

18 January 2023 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
HZHao Zhang
Hao Zhang
THTing-Zhu Huang
Ting-Zhu Huang
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
Wei HeWei He
Wei He
JCJae Kyu Choi
Jae Kyu Choi
Yu-Bang ZhengYu-Bang Zheng
Yu-Bang Zheng
Recently, the transform-based tensor nuclear norm (TNN) framework has yielded promising results for hyperspectral image (HSI) denoising as compared with previous original-domain tensor-based models. However, the TNN framework only exploits the low-rankness of each band of HSIs (tensors) under a single spectral transform. The correlation between all bands under the transform (i.e., the global low-rankness of the transformed tensor) and the sparsity of the transformed HSI, which are beneficial for HSI denoising, are usually neglected in the TNN framework. In this paper, we propose to reconcile sparse and low-tensor-ring (TR)-rank priors in the learned transformed domain (called T-RSTR model) for HSI denoising. In T-RSTR, the transform-based low-TR-rank and sparse regularizers are designed to characterize the global low-rankness and sparsity of the transformed tensors respectively, and then the transform-based low-TR-rank and sparse regularizers are organically integrated and benefit from each other for substantially boosting denoising performance. To tackle the T-RSTR model, we elaborately design a proximal alternating minimization-based algorithm with the theoretical convergence. Extensive numerical results demonstrate that T-RSTR is superior to the competing methods.

29 September 2022 in Journal of Scientific Computing
Verified
Journal Article
Closed access
PDF file
CLCheng-Yao Lyu
Cheng-Yao Lyu
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
BLBen-Zheng Li
Ben-Zheng Li
HZHao Zhang
Hao Zhang
THTing-Zhu Huang
Ting-Zhu Huang

13 December 2023 in IEEE Transactions on Big Data
Verified
Journal Article
Closed access
ZHZhi-Long Han
Zhi-Long Han
THTing-Zhu Huang
Ting-Zhu Huang
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
HZHao Zhang
Hao Zhang
YLYun-Yang Liu
Yun-Yang Liu
Multi-dimensional data are inevitably corrupted, which hinders subsequent applications (e.g., image segmentation and classification). Recently, due to the powerful ability to characterize the correlation between any two modes of tensors, fully-connected tensor network (FCTN) decomposition has received increasing attention in multi-dimensional data recovery. However, the expressive power of FCTN decomposition in the original pixel domain has yet to be fully leveraged, which can not provide satisfactory results in the recovery of details and textures, especially for low-sampling rates or heavy noise scenarios. In this work, we suggest a feature-based FCTN decomposition model (termed as F-FCTN) for multi-dimensional data recovery, which can faithfully capture the relationship between the spatial-temporal/spectral-feature modes. Compared with the original FCTN decomposition, F-FCTN can more effectively recover the details and textures and be more suitable for the subsequent high-level applications. However, F-FCTN leads to a larger-scale feature tensor as compared with the original tensor, which brings challenges in designing the solving algorithm. To harness the resulting large-scale optimization problem, we develop an efficient leverage score sampling-based proximal alternating minimization (S-PAM) algorithm and theoretically establish its relative error guarantee. Extensive numerical experiments on real-world data illustrate that the proposed method performs favorably against compared methods in data recovery and facilitates subsequent image classification.

14 July 2025 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
SLSheng Liu
Sheng Liu
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
HZHao Zhang
Hao Zhang
Recently, tensor decompositions have received significant attention for processing multi-dimensional signals, especially representative by the block-term decomposition (BTD) family and the tensor network decomposition (TND) family. However, these two families have long been isolated from each other, with their respective wisdom neither inspiring nor benefiting each other. To address this dilemma, we propose a block tensor ring decomposition (BTRD), which decomposes an Nth-order tensor into a sum of outer products between basic vector factors and the (N − 1)th-order coefficient tensors, which are further represented using a tensor ring. The benefit of the BTRD is that it can better explore outer multiple components structure of the tensor and inner tensor topology of each component. To examine the potential of the proposed BTRD, we apply it to a low-rank tensor completion model as a representative task and prove a theoretical generalization error bound which provides a theoretical perspective to support the advantages of the proposed model for higher-order tensors. To address the resulting optimization problem, we apply an efficient proximal alternating minimization (PAM)-based algorithm with a theoretical convergence guarantee. Extensive experimental results on real-world signal data (color videos and light field images) demonstrate the superiority of the proposed model against the state-of-the-art baseline models.

19 October 2023 in IEEE Signal Processing Letters
Verified
Journal Article
Closed access
WWWei-Hao Wu
Wei-Hao Wu
THTing-Zhu Huang
Ting-Zhu Huang
HZHao Zhang
Hao Zhang
JWJian-Li Wang
Jian-Li Wang
Xi-Le ZhaoXi-Le Zhao
Xi-Le Zhao
Recently, untrained neural network priors (UNNPs) have received increasing attention for multi-dimensional image recovery. However, previous studies are based on over-parameterized untrained neural networks, which results in unstable behavior. In this letter, we propose an untrained low-rank neural network prior (ULRNNP) for multi-dimensional image recovery, which enjoys the powerful representation ability and stable behavior. More specifically, the elaborately designed nonlinear Tucker decomposition module implicitly imposes low-rank constraints on the feature tensor and can more compactly represent the feature tensor. Attributed to the suggested nonlinear Tucker decomposition module, ULRNNP can simultaneously enjoy strong representation ability and stable behavior. The friendly stable behavior allows us to design a friendly stopping criteria without the reference ground truth image as compared with classic UNNP-based methods. Extensive experiments on different multi-dimensional image datasets validate the superior performance of the proposed ULRNNP over state-of-the-art methods.