Qian Zhao
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
Research Interests
Qian Zhao has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Dongming Zhou
School of Information Science and Engineering, Yunnan University, Kunming, China
4 shared publications - Hao Yang
School of Information Science and Engineering, Yunnan University, Kunming, China
3 shared publications - Gang Li
Shanghai Research Institute for Autonomous Intelligent Systems, Tongji University, Shanghai, China
2 shared publications - Jinde Cao
School of Mathematics, Southeast University, Nanjing 210096, People’s Republic of China
2 shared publications - Bin He
Tongji University
1 shared publication - Changcheng Wang
School of Information and Engineering, Yunnan University, Kunming, China
1 shared publication
Latest Publications
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 28, 11889-11902
Illumination-Invariant Representation Learning With Multiple Priors for Low-Light Object Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 28, 11889-11902
Illumination-Invariant Representation Learning With Multiple Priors for Low-Light Object Detection
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. PP (2162237X) , 1-21
Deep Learning for Low-Light Vision: A Comprehensive Survey
Visual recognition in low-light environments is a challenging problem since degraded images are the stacking of multiple degradations (noise, low light and blur, etc.). It has received extensive attention from academia and industry in the era of deep learning. Existing surveys focus on low-light image enhancement (LLIE) methods and normal-light visual recognition methods, while few comprehensive surveys of low-light-related vision tasks. This article provides a comprehensive survey of the latest advancements in low-light vision, including methods, datasets, and evaluation metrics, in two aspects: visual quality-driven and recognition quality-driven. On the visual quality-driven aspect, we survey a large number of very recent LLIE methods. On the recognition quality-driven aspect, we survey low-light object detection techniques in the deep learning era using more intuitive categorization method. Furthermore, a quantitative benchmarking of different methods is conducted on several widely adopted low-light vision-related datasets. Finally, we discuss the challenges that exist in low-light vision and future directions worth exploring. We provide a public website that will continue to track developments in this promising field.
- 2 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. PP (2162237X) , 1-21
Deep Learning for Low-Light Vision: A Comprehensive Survey
Visual recognition in low-light environments is a challenging problem since degraded images are the stacking of multiple degradations (noise, low light and blur, etc.). It has received extensive attention from academia and industry in the era of deep learning. Existing surveys focus on low-light image enhancement (LLIE) methods and normal-light visual recognition methods, while few comprehensive surveys of low-light-related vision tasks. This article provides a comprehensive survey of the latest advancements in low-light vision, including methods, datasets, and evaluation metrics, in two aspects: visual quality-driven and recognition quality-driven. On the visual quality-driven aspect, we survey a large number of very recent LLIE methods. On the recognition quality-driven aspect, we survey low-light object detection techniques in the deep learning era using more intuitive categorization method. Furthermore, a quantitative benchmarking of different methods is conducted on several widely adopted low-light vision-related datasets. Finally, we discuss the challenges that exist in low-light vision and future directions worth exploring. We provide a public website that will continue to track developments in this promising field.
- Open Access
- 24 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1098-1104
Research based on Mixed Sample Data Augmentation method
Mixed Sample Data Augmentation (MSDA) has flourished in recent years, with many successful methods. These approaches have greatly alleviated neural network problems due to overfitting, class imbalance, adversarial attacks, etc. However, most of these methods fail to distinguish between image style and content, which causes the network to be biased to focus on a small subset of information to learn the features of the category. To further improve the performance of the model, this paper proposes a new method based on MSDA. This approach not only generates samples that contain more information at larger spatial scales, but also enables the network to manipulate the content and style information of input image pairs separately by generating samples that transfer their styles. Through extensive experiments, we compare the performance of previous approaches, and this approach has been shown to improve the classification accuracy of the model and make the model more general and robust.
- 24 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 1098-1104
Open AccessResearch based on Mixed Sample Data Augmentation method
Mixed Sample Data Augmentation (MSDA) has flourished in recent years, with many successful methods. These approaches have greatly alleviated neural network problems due to overfitting, class imbalance, adversarial attacks, etc. However, most of these methods fail to distinguish between image style and content, which causes the network to be biased to focus on a small subset of information to learn the features of the category. To further improve the performance of the model, this paper proposes a new method based on MSDA. This approach not only generates samples that contain more information at larger spatial scales, but also enables the network to manipulate the content and style information of input image pairs separately by generating samples that transfer their styles. Through extensive experiments, we compare the performance of previous approaches, and this approach has been shown to improve the classification accuracy of the model and make the model more general and robust.
- 19 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-15
Rethinking Image Deblurring via CNN-Transformer Multiscale Hybrid Architecture
Image deblurring is a representative low-level vision task that aims to estimate latent sharp images from blurred images. Recently, convolutional neural network (CNN)-based methods have dominated image deblurring. However, traditional CNN-based deblurring methods suffer from two essential issues: first, existing multiscale deblurring methods process blurred images at different scales through sub-networks with the same composition, which limits the model performance. Second, the convolutional layers fail to adapt to the input content and cannot effectively capture long-range dependencies. To alleviate the above issues, we rethink the multiscale architecture that follows a coarse-to-fine strategy and propose a novel hybrid architecture that combines CNN and transformer (CTMS). CTMS has three distinct features. First, the finer-scale sub-networks in CTMS are designed as architectures with larger receptive fields to obtain the pixel values around the blur, which can be used to efficiently handle large-area blur. Then, we propose a feature modulation network to alleviate the disadvantages of CNN sub-networks that lack input content adaptation. Finally, we design an efficient transformer block, which significantly reduces the computational burden and requires no pre-training. Our proposed deblurring model is extensively evaluated on several benchmark datasets, and achieves superior performance compared to state-of-the-art deblurring methods. Especially, the peak signal to noise ratio (PSNR) and structural similarity (SSIM) values are 32.73 dB and 0.959, respectively, on the popular dataset GoPro. In addition, we conduct joint evaluation experiments on the proposed method deblurring performance, object detection, and image segmentation to demonstrate the effectiveness of CTMS for subsequent high-level computer vision tasks.
- 19 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 72 (00189456) , 1-15
Rethinking Image Deblurring via CNN-Transformer Multiscale Hybrid Architecture
Image deblurring is a representative low-level vision task that aims to estimate latent sharp images from blurred images. Recently, convolutional neural network (CNN)-based methods have dominated image deblurring. However, traditional CNN-based deblurring methods suffer from two essential issues: first, existing multiscale deblurring methods process blurred images at different scales through sub-networks with the same composition, which limits the model performance. Second, the convolutional layers fail to adapt to the input content and cannot effectively capture long-range dependencies. To alleviate the above issues, we rethink the multiscale architecture that follows a coarse-to-fine strategy and propose a novel hybrid architecture that combines CNN and transformer (CTMS). CTMS has three distinct features. First, the finer-scale sub-networks in CTMS are designed as architectures with larger receptive fields to obtain the pixel values around the blur, which can be used to efficiently handle large-area blur. Then, we propose a feature modulation network to alleviate the disadvantages of CNN sub-networks that lack input content adaptation. Finally, we design an efficient transformer block, which significantly reduces the computational burden and requires no pre-training. Our proposed deblurring model is extensively evaluated on several benchmark datasets, and achieves superior performance compared to state-of-the-art deblurring methods. Especially, the peak signal to noise ratio (PSNR) and structural similarity (SSIM) values are 32.73 dB and 0.959, respectively, on the popular dataset GoPro. In addition, we conduct joint evaluation experiments on the proposed method deblurring performance, object detection, and image segmentation to demonstrate the effectiveness of CTMS for subsequent high-level computer vision tasks.
- 31 October 2022
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 79 (6) , 6115-6140
RainFormer: a pyramid transformer for single image deraining
- 31 October 2022
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 79 (6) , 6115-6140
RainFormer: a pyramid transformer for single image deraining
- 29 July 2022
- Published by Springer Nature in Neural Processing Letters
- Vol. 55 (4) , 3985-4006
CDMC-Net: Context-Aware Image Deblurring Using a Multi-scale Cascaded Network
- 29 July 2022
- Published by Springer Nature in Neural Processing Letters
- Vol. 55 (4) , 3985-4006
CDMC-Net: Context-Aware Image Deblurring Using a Multi-scale Cascaded Network