Zhihua Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Zhihua Chen has a total of 80 co-authors. Here are the top co-authors by number of shared publications:
- Lei Dai
Department of Computer Science and Engineering, East China University of Science and Technology, Shanghai, China
19 shared publications - Bin Sheng
Shanghai Jiao Tong University
12 shared publications - Ping Li
The Hong Kong Polytechnic University
9 shared publications Bin Sheng
4 shared publicationsPing Li
4 shared publications- Jingxuan Zhang
Department of Computer Science and Engineering, East China University of Science and Technology, Shanghai, China
3 shared publications
Latest Publications
- 8 July 2026
- Published by Springer Nature in The Visual Computer
- Vol. 42  (9) , 392
TDNet: degradation-aware comprehensive task decomposition for joint rain and haze removal
- 8 July 2026
- Published by Springer Nature in The Visual Computer
- Vol. 42  (9) , 392
TDNet: degradation-aware comprehensive task decomposition for joint rain and haze removal
- 20 May 2026
- Published by Springer Nature in The Visual Computer
- Vol. 42  (8) , 306
CCBNet: collaborative calibration and bridging network for visible-infrared person re-identification
- 20 May 2026
- Published by Springer Nature in The Visual Computer
- Vol. 42  (8) , 306
CCBNet: collaborative calibration and bridging network for visible-infrared person re-identification
- Top Cited Papers
- 6 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP  (01628828) , 1-18
Dataset Distillation via a Noise-Unconstrained Generative Model
Dataset distillation (DD) aims to synthesize a more compact dataset than the original one and models trained on it are expected to have the same generalization capabilities as on the original dataset. Previous work via a generative model (GM) faces several limitations. First, GM struggles to generate representative samples due to a lack of constraints. Second, it overlooks the relationships between generated samples, limiting its effectiveness. In this paper, a new noise-unconstrained GM-based DD framework is proposed. In the distillation stage, an adaptive matching coefficient is introduced to align generated images with representative class elements and the MiniMax loss function is extended to reduce the optimization difficulty. In the deployment stage, features among each generative image are ensembled by gradient-matching based DD. Theoretical analysis based on McDiarmid's inequality demonstrates that the proposed components can reduce the generalization error of the original baseline method. We also provide insights into the potential of generated images as an effective proxy dataset for DD. For example, on the ImageWoof dataset with 50 distilled images per class using a 6-layer ConvNet for evaluation, generated images outperform 25%, 50%, and 75% original images by 8.4%, 6.3%, and 8.3% in distillation performance. Our method effectively handles both low- and high-resolution datasets, with experiments on 11 benchmarks demonstrating its efficacy.
- 6 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP  (01628828) , 1-18
Top Cited PapersDataset Distillation via a Noise-Unconstrained Generative Model
Dataset distillation (DD) aims to synthesize a more compact dataset than the original one and models trained on it are expected to have the same generalization capabilities as on the original dataset. Previous work via a generative model (GM) faces several limitations. First, GM struggles to generate representative samples due to a lack of constraints. Second, it overlooks the relationships between generated samples, limiting its effectiveness. In this paper, a new noise-unconstrained GM-based DD framework is proposed. In the distillation stage, an adaptive matching coefficient is introduced to align generated images with representative class elements and the MiniMax loss function is extended to reduce the optimization difficulty. In the deployment stage, features among each generative image are ensembled by gradient-matching based DD. Theoretical analysis based on McDiarmid's inequality demonstrates that the proposed components can reduce the generalization error of the original baseline method. We also provide insights into the potential of generated images as an effective proxy dataset for DD. For example, on the ImageWoof dataset with 50 distilled images per class using a 6-layer ConvNet for evaluation, generated images outperform 25%, 50%, and 75% original images by 8.4%, 6.3%, and 8.3% in distillation performance. Our method effectively handles both low- and high-resolution datasets, with experiments on 11 benchmarks demonstrating its efficacy.
- Top Cited Papers
- 18 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36  (6) , 7540-7553
LODNeuS: A Flexible Lightweight Neural Implicit Surface Representation With Unconstrained Viewpoint Rendering
NeRF-like methods learn implicit 3D neural representations from 2D multiview images, enabling the synthesis of compelling novel views. However, to capture high-fidelity geometry, prior methods often rely on large-scale networks. This dependency hampers the potential applications of neural implicit representations, such as MR visualization. To address this, we introduce LODNeuS, an implicit surface representation based on feature voxel grids. LODNeuS captures multiple LODs of implicit geometry by maintaining voxel grids paired with a set of corresponding lightweight decoders. This allows for high-quality rendering with the ability to dynamically switch between detail levels. Another challenge is that existing methods, both volumetric and surface-based, tend to train and render their representations within a confined space, without explicitly restricting the sampling points properly. This lack of constraints can result in ambiguity, artifacts, and inefficient use of computational resources. We study this effect during free viewpoint rendering using conventional methods and develop an adaptive sampling scheme that emphasizes a valid geometric space for sampling point allocation. Our experimental results show that LODNeuS can match the visual quality of existing methods while offering flexible and lightweight inference. The benefits of adaptive sampling are also demonstrated in the free viewpoint rendering subsection. Our work extends the capabilities of neural implicit representations beyond previously defined limitations, broadening the scope of potential applications.
- 18 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36  (6) , 7540-7553
Top Cited PapersLODNeuS: A Flexible Lightweight Neural Implicit Surface Representation With Unconstrained Viewpoint Rendering
NeRF-like methods learn implicit 3D neural representations from 2D multiview images, enabling the synthesis of compelling novel views. However, to capture high-fidelity geometry, prior methods often rely on large-scale networks. This dependency hampers the potential applications of neural implicit representations, such as MR visualization. To address this, we introduce LODNeuS, an implicit surface representation based on feature voxel grids. LODNeuS captures multiple LODs of implicit geometry by maintaining voxel grids paired with a set of corresponding lightweight decoders. This allows for high-quality rendering with the ability to dynamically switch between detail levels. Another challenge is that existing methods, both volumetric and surface-based, tend to train and render their representations within a confined space, without explicitly restricting the sampling points properly. This lack of constraints can result in ambiguity, artifacts, and inefficient use of computational resources. We study this effect during free viewpoint rendering using conventional methods and develop an adaptive sampling scheme that emphasizes a valid geometric space for sampling point allocation. Our experimental results show that LODNeuS can match the visual quality of existing methods while offering flexible and lightweight inference. The benefits of adaptive sampling are also demonstrated in the free viewpoint rendering subsection. Our work extends the capabilities of neural implicit representations beyond previously defined limitations, broadening the scope of potential applications.
- 29 January 2026
- Published by Elsevier in Neural Networks
- Vol. 199, 108657
Autorep: Automatic network search with structured reparameterized based linear operation expansion and gradient proxy guided reduction
- 29 January 2026
- Published by Elsevier in Neural Networks
- Vol. 199, 108657
Autorep: Automatic network search with structured reparameterized based linear operation expansion and gradient proxy guided reduction
- 26 December 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 166
Slimmable neural architecture design based on cross architecture and token distillation
- 26 December 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 166
Slimmable neural architecture design based on cross architecture and token distillation
- 5 December 2025
- Published by Springer Nature in The Visual Computer
- Vol. 42  (1) , 11
Robust piglet counting in crowded environments via small trajectory fusion and multi-line counting module
- 5 December 2025
- Published by Springer Nature in The Visual Computer
- Vol. 42  (1) , 11
Robust piglet counting in crowded environments via small trajectory fusion and multi-line counting module
- 28 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 34  (10577149) , 7050-7064
Unleashing the Power of Each Distilled Image
Dataset distillation (DD) aims to accelerate the training speed of neural networks (NNs) by synthesizing a reduced dataset. NNs trained on the smaller dataset are expected to obtain almost the same test set accuracy as they do on the larger one. Previous DD research treated the obtained distilled dataset as a regular dataset for training, neglecting the overfitting issue caused by the limited number of original distilled images. In this paper, we propose a new DD paradigm. Specifically, in the deployment stage, distilled images are augmented by amplifying their local information since the teacher network can produce diverse supervision signals when receiving inputs from different regions. Efficient and diverse augmentation methods for each distilled image are devised, while ensuring the authenticity of augmented samples. Additionally, to alleviate the increased training cost caused by data augmentation, we design a bi-directional dynamic dataset pruning technique to prune the original distilled dataset and augmented distilled dataset. A new pruning strategy and scheduling are proposed based on experimental findings. Experiments on 9 benchmark datasets (CIFAR10, CIFAR100, ImageWoof, ImageCat, ImageFruit, ImageNette, ImageNet10, ImageNet100 and ImageNet1K) demonstrate the effectiveness of our approach. For instance, on the ImageNet1K dataset with a ResNet18 architecture and 50 distilled images per class, our algorithm surpasses the second-ranked MiniMax algorithm by 7.6%, achieving a distilled accuracy of 66.2%.
- 28 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 34  (10577149) , 7050-7064
Unleashing the Power of Each Distilled Image
Dataset distillation (DD) aims to accelerate the training speed of neural networks (NNs) by synthesizing a reduced dataset. NNs trained on the smaller dataset are expected to obtain almost the same test set accuracy as they do on the larger one. Previous DD research treated the obtained distilled dataset as a regular dataset for training, neglecting the overfitting issue caused by the limited number of original distilled images. In this paper, we propose a new DD paradigm. Specifically, in the deployment stage, distilled images are augmented by amplifying their local information since the teacher network can produce diverse supervision signals when receiving inputs from different regions. Efficient and diverse augmentation methods for each distilled image are devised, while ensuring the authenticity of augmented samples. Additionally, to alleviate the increased training cost caused by data augmentation, we design a bi-directional dynamic dataset pruning technique to prune the original distilled dataset and augmented distilled dataset. A new pruning strategy and scheduling are proposed based on experimental findings. Experiments on 9 benchmark datasets (CIFAR10, CIFAR100, ImageWoof, ImageCat, ImageFruit, ImageNette, ImageNet10, ImageNet100 and ImageNet1K) demonstrate the effectiveness of our approach. For instance, on the ImageNet1K dataset with a ResNet18 architecture and 50 distilled images per class, our algorithm surpasses the second-ranked MiniMax algorithm by 7.6%, achieving a distilled accuracy of 66.2%.
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74  (00189456) , 1-15
A Two-Stage Exposure Controllable Network for Image Enhancement With Low-Light Degradation
Controllable low-light enhancement is beneficial to both image aesthetics and high-level vision tasks. Existing methods fail to generate smooth brightness transitions with low cost. In this paper, we propose an unsupervised two-stage exposure controllable network (TSECN) featuring low-cost and natural brightness customization. Conditioning on a color-correlation based exposure reference map (CERM), an offline naturalness-preserving exposure controller (ONEC) is implemented by modulating a latent scene illumination representation, which is predicted via a two-stage progressive enhancement pipeline (TPE). Specifically, the first stage performs unsupervised low-light denoising and improves visibility with deep gamma correction. The latter handles fine-grained quality refinement and representation learning. Extensive evaluations demonstrate the superior performance of our method across eight real-world datasets. On three paired datasets, TSECN outperforms advanced methods by an average PSNR improvement of 2.29 dB. Experiments on two high-level vision tasks also validate its value in application-oriented scenarios. The code will be available at https://github.com/cherrysherryplus/TSECN.
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74  (00189456) , 1-15
A Two-Stage Exposure Controllable Network for Image Enhancement With Low-Light Degradation
Controllable low-light enhancement is beneficial to both image aesthetics and high-level vision tasks. Existing methods fail to generate smooth brightness transitions with low cost. In this paper, we propose an unsupervised two-stage exposure controllable network (TSECN) featuring low-cost and natural brightness customization. Conditioning on a color-correlation based exposure reference map (CERM), an offline naturalness-preserving exposure controller (ONEC) is implemented by modulating a latent scene illumination representation, which is predicted via a two-stage progressive enhancement pipeline (TPE). Specifically, the first stage performs unsupervised low-light denoising and improves visibility with deep gamma correction. The latter handles fine-grained quality refinement and representation learning. Extensive evaluations demonstrate the superior performance of our method across eight real-world datasets. On three paired datasets, TSECN outperforms advanced methods by an average PSNR improvement of 2.29 dB. Experiments on two high-level vision tasks also validate its value in application-oriented scenarios. The code will be available at https://github.com/cherrysherryplus/TSECN.
- 5 July 2025
- Published by Elsevier in Neural Networks
- Vol. 191, 107819
Gradient amplification for gradient matching based dataset distillation
- 5 July 2025
- Published by Elsevier in Neural Networks
- Vol. 191, 107819
Gradient amplification for gradient matching based dataset distillation