Chongzhi Zhang
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
Research Subjects
Chongzhi Zhang has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
- Aishan Liu
Beihang University, Beijing, China
3 shared publications - Hang Yu
State Key Laboratory of Software Development Environment, Beihang University, Beijing, China
3 shared publications - Aixin Sun
Nanyang technological university, Singapore, Singapore
2 shared publications - Xianglong Liu
Beihang University, Beijing, China
2 shared publications - Xizhou Zhu
SenseTime Research, Beijing, China
2 shared publications - Changjin Ao
Key Laboratory of Animal Feed and Nutrition of Inner Mongolia Autonomous Region, College of Animal Science, Inner Mongolia Agricultural University, Hohhot 010018, China
1 shared publication
Latest Publications
- Open Access
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 313-324
A Flexible and Scalable Framework for Video Moment Search
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 313-324
Open AccessA Flexible and Scalable Framework for Video Moment Search
- Open Access
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 231-239
TVR-Ranking: A Dataset for Ranked Video Moment Retrieval with Imprecise Queries
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 231-239
Open AccessTVR-Ranking: A Dataset for Ranked Video Moment Retrieval with Imprecise Queries
- 26 October 2024
- Published by Springer Nature
- p. 397-421
Large Motion Model for Unified Multi-modal Motion Generation
- 26 October 2024
- Published by Springer Nature
- p. 397-421
Large Motion Model for Unified Multi-modal Motion Generation
- Open Access
- 5 October 2023
- Published by IOS Press
- p. 89-94
Construction of Single Evaporative Emission Model and Analysis of Emission Characteristics for Gasoline Vehicles
To calculate the proportion of evaporative emissions of the China VI, this paper establishes a single-vehicle evaporative emission model for light gasoline vehicles, and an activity level model based on the car use habits of the employees of China Automotive Technology Research Center Co., Ltd. The proportion of evaporative emissions from Vehicle-A and Vehicle B, the running loss emissions of both Vehicle-A and Vehicle B account for over 50%, while refueling emissions and the diurnal 48-hour emissions account for lower proportions. The total annual evaporative emissions of vehicle A can reach 353.496 g, and the total annual evaporative emissions of vehicle B can reach 589.758 g.
- 5 October 2023
- Published by IOS Press
- p. 89-94
Open AccessConstruction of Single Evaporative Emission Model and Analysis of Emission Characteristics for Gasoline Vehicles
To calculate the proportion of evaporative emissions of the China VI, this paper establishes a single-vehicle evaporative emission model for light gasoline vehicles, and an activity level model based on the car use habits of the employees of China Automotive Technology Research Center Co., Ltd. The proportion of evaporative emissions from Vehicle-A and Vehicle B, the running loss emissions of both Vehicle-A and Vehicle B account for over 50%, while refueling emissions and the diurnal 48-hour emissions account for lower proportions. The total annual evaporative emissions of vehicle A can reach 353.496 g, and the total annual evaporative emissions of vehicle B can reach 589.758 g.
- 26 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 8955-8967
Progressive Diversified Augmentation for General Robustness of DNNs: A Unified Approach
Adversarial images are imperceptible perturbations to mislead deep neural networks (DNNs), which have attracted great attention in recent years. Although several defense strategies achieved encouraging robustness against adversarial samples, most of them still failed to consider the robustness on common corruptions (e.g. noise, blur, and weather/digital effects). To address this problem, we propose a simple yet effective method, named Progressive Diversified Augmentation (PDA), which improves the robustness of DNNs by progressively injecting diverse adversarial noises during training. In other words, DNNs trained with PDA achieve better general robustness against both adversarial attacks and common corruptions than other strategies. In addition, PDA also enjoys the advantages of spending less training time and keeping high standard accuracy on clean examples. Further, we theoretically prove that PDA can control the perturbation bound and guarantee better robustness. Extensive results on CIFAR-10, SVHN, ImageNet, CIFAR-10-C and ImageNet-C have demonstrated that PDA comprehensively outperforms its counterparts on the robustness against adversarial examples and common corruptions as well as clean images. More experiments on the frequency-based perturbations and visualized gradients further prove that PDA achieves general robustness and is more aligned with the human visual system.
- 26 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 8955-8967
Progressive Diversified Augmentation for General Robustness of DNNs: A Unified Approach
Adversarial images are imperceptible perturbations to mislead deep neural networks (DNNs), which have attracted great attention in recent years. Although several defense strategies achieved encouraging robustness against adversarial samples, most of them still failed to consider the robustness on common corruptions (e.g. noise, blur, and weather/digital effects). To address this problem, we propose a simple yet effective method, named Progressive Diversified Augmentation (PDA), which improves the robustness of DNNs by progressively injecting diverse adversarial noises during training. In other words, DNNs trained with PDA achieve better general robustness against both adversarial attacks and common corruptions than other strategies. In addition, PDA also enjoys the advantages of spending less training time and keeping high standard accuracy on clean examples. Further, we theoretically prove that PDA can control the perturbation bound and guarantee better robustness. Extensive results on CIFAR-10, SVHN, ImageNet, CIFAR-10-C and ImageNet-C have demonstrated that PDA comprehensively outperforms its counterparts on the robustness against adversarial examples and common corruptions as well as clean images. More experiments on the frequency-based perturbations and visualized gradients further prove that PDA achieves general robustness and is more aligned with the human visual system.
- 23 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 5769-5781
Training Robust Deep Neural Networks via Adversarial Noise Propagation
In practice, deep neural networks have been found to be vulnerable to various types of noise, such as adversarial examples and corruption. Various adversarial defense methods have accordingly been developed to improve adversarial robustness for deep models. However, simply training on data mixed with adversarial examples, most of these models still fail to defend against the generalized types of noise. Motivated by the fact that hidden layers play a highly important role in maintaining a robust model, this paper proposes a simple yet powerful training algorithm, named Adversarial Noise Propagation (ANP), which injects noise into the hidden layers in a layer-wise manner. ANP can be implemented efficiently by exploiting the nature of the backward-forward training style. Through thorough investigations, we determine that different hidden layers make different contributions to model robustness and clean accuracy, while shallow layers are comparatively more critical than deep layers. Moreover, our framework can be easily combined with other adversarial training methods to further improve model robustness by exploiting the potential of hidden layers. Extensive experiments on MNIST, CIFAR-10, CIFAR-10-C, CIFAR-10-P, and ImageNet demonstrate that ANP enables the strong robustness for deep models against both adversarial and corrupted ones, and also significantly outperforms various adversarial defense methods.
- 23 June 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 5769-5781
Training Robust Deep Neural Networks via Adversarial Noise Propagation
In practice, deep neural networks have been found to be vulnerable to various types of noise, such as adversarial examples and corruption. Various adversarial defense methods have accordingly been developed to improve adversarial robustness for deep models. However, simply training on data mixed with adversarial examples, most of these models still fail to defend against the generalized types of noise. Motivated by the fact that hidden layers play a highly important role in maintaining a robust model, this paper proposes a simple yet powerful training algorithm, named Adversarial Noise Propagation (ANP), which injects noise into the hidden layers in a layer-wise manner. ANP can be implemented efficiently by exploiting the nature of the backward-forward training style. Through thorough investigations, we determine that different hidden layers make different contributions to model robustness and clean accuracy, while shallow layers are comparatively more critical than deep layers. Moreover, our framework can be easily combined with other adversarial training methods to further improve model robustness by exploiting the potential of hidden layers. Extensive experiments on MNIST, CIFAR-10, CIFAR-10-C, CIFAR-10-P, and ImageNet demonstrate that ANP enables the strong robustness for deep models against both adversarial and corrupted ones, and also significantly outperforms various adversarial defense methods.
- 8 December 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 1291-1304
Interpreting and Improving Adversarial Robustness of Deep Neural Networks With Neuron Sensitivity
Deep neural networks (DNNs) are vulnerable to adversarial examples where inputs with imperceptible perturbations mislead DNNs to incorrect results. Despite the potential risk they bring, adversarial examples are also valuable for providing insights into the weakness and blind-spots of DNNs. Thus, the interpretability of a DNN in the adversarial setting aims to explain the rationale behind its decision-making process and makes deeper understanding which results in better practical applications. To address this issue, we try to explain adversarial robustness for deep models from a new perspective of neuron sensitivity which is measured by neuron behavior variation intensity against benign and adversarial examples. In this paper, we first draw the close connection between adversarial robustness and neuron sensitivities, as sensitive neurons make the most non-trivial contributions to model predictions in the adversarial setting. Based on that, we further propose to improve adversarial robustness by stabilizing the behaviors of sensitive neurons. Moreover, we demonstrate that state-of-the-art adversarial training methods improve model robustness by reducing neuron sensitivities, which in turn confirms the strong connections between adversarial robustness and neuron sensitivity. Extensive experiments on various datasets demonstrate that our algorithm effectively achieves excellent results. To the best of our knowledge, we are the first to study adversarial robustness using neuron sensitivities.
- 8 December 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 1291-1304
Interpreting and Improving Adversarial Robustness of Deep Neural Networks With Neuron Sensitivity
Deep neural networks (DNNs) are vulnerable to adversarial examples where inputs with imperceptible perturbations mislead DNNs to incorrect results. Despite the potential risk they bring, adversarial examples are also valuable for providing insights into the weakness and blind-spots of DNNs. Thus, the interpretability of a DNN in the adversarial setting aims to explain the rationale behind its decision-making process and makes deeper understanding which results in better practical applications. To address this issue, we try to explain adversarial robustness for deep models from a new perspective of neuron sensitivity which is measured by neuron behavior variation intensity against benign and adversarial examples. In this paper, we first draw the close connection between adversarial robustness and neuron sensitivities, as sensitive neurons make the most non-trivial contributions to model predictions in the adversarial setting. Based on that, we further propose to improve adversarial robustness by stabilizing the behaviors of sensitive neurons. Moreover, we demonstrate that state-of-the-art adversarial training methods improve model robustness by reducing neuron sensitivities, which in turn confirms the strong connections between adversarial robustness and neuron sensitivity. Extensive experiments on various datasets demonstrate that our algorithm effectively achieves excellent results. To the best of our knowledge, we are the first to study adversarial robustness using neuron sensitivities.
- 1 June 2014
- Published by Elsevier in Animal Reproduction Science
- Vol. 147 (3-4) , 99-105
Effects of maternal undernutrition during late pregnancy on the development and function of ovine fetal liver
- 1 June 2014
- Published by Elsevier in Animal Reproduction Science
- Vol. 147 (3-4) , 99-105
Effects of maternal undernutrition during late pregnancy on the development and function of ovine fetal liver