Xiangrong Zeng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiangrong Zeng has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
Maojun Zhang
College of Systems Engineering National University of Defense Technology Changsha 410073 China
2 shared publicationsZhiwei Zhong
College of Systems Engineering National University of Defense Technology Changsha 410073 China
2 shared publications- Changhua Jiang
National Key Laboratory of Human Factors Engineering, China Astronaut Research and Training Center, Beijing, China
1 shared publication - Maojun Zhang
National University of Defense Technology
1 shared publication - Sheng Tan
DCapture Group, FDVUST, China
1 shared publication Yan Liu
College of Systems Engineering National University of Defense Technology Changsha 410073 China
1 shared publication
Latest Publications
- Open Access
- 1 February 2023
- Published by MDPI AG in Electronics
- Vol. 12 (3) , 732
Energy-Constrained Deep Neural Network Compression for Depth Estimation
Many applications, such as autonomous driving, robotics, etc., require accurately estimating depth in real time. Currently, deep learning is the most popular approach to stereo depth estimation. Some of these models have to operate in highly energy-constrained environments, while they are usually computationally intensive, containing massive parameter sets ranging from thousands to millions. This makes them hard to perform on low-power devices with limited storage in practice. To overcome this shortcoming, we model the training process of a deep neural network (DNN) for depth estimation under a given energy constraint as a constrained optimization problem and solve it through a proposed projected adaptive cubic quasi-Newton method (termed ProjACQN). Moreover, the trained model is also deployed on a GPU and an embedded device to evaluate its performance. Experiments show that the stage four results of ProjACQN on the KITTI-2012 and KITTI-2015 datasets under a 70% energy budget achieve (1) 0.13% and 0.61%, respectively, lower three-pixel error than the state-of-the-art ProjAdam when put on a single RTX 3090Ti; (2) 4.82% and 7.58%, respectively, lower three-pixel error than the pruning method Lottery-Ticket; (3) 5.80% and 0.12%, respectively, lower three-pixel error than ProjAdam on the embedded device Nvidia Jetson AGX Xavier. These results show that our method can reduce the energy consumption of depth estimation DNNs while maintaining their accuracy.
- 1 February 2023
- Published by MDPI AG in Electronics
- Vol. 12 (3) , 732
Open AccessEnergy-Constrained Deep Neural Network Compression for Depth Estimation
Many applications, such as autonomous driving, robotics, etc., require accurately estimating depth in real time. Currently, deep learning is the most popular approach to stereo depth estimation. Some of these models have to operate in highly energy-constrained environments, while they are usually computationally intensive, containing massive parameter sets ranging from thousands to millions. This makes them hard to perform on low-power devices with limited storage in practice. To overcome this shortcoming, we model the training process of a deep neural network (DNN) for depth estimation under a given energy constraint as a constrained optimization problem and solve it through a proposed projected adaptive cubic quasi-Newton method (termed ProjACQN). Moreover, the trained model is also deployed on a GPU and an embedded device to evaluate its performance. Experiments show that the stage four results of ProjACQN on the KITTI-2012 and KITTI-2015 datasets under a 70% energy budget achieve (1) 0.13% and 0.61%, respectively, lower three-pixel error than the state-of-the-art ProjAdam when put on a single RTX 3090Ti; (2) 4.82% and 7.58%, respectively, lower three-pixel error than the pruning method Lottery-Ticket; (3) 5.80% and 0.12%, respectively, lower three-pixel error than ProjAdam on the embedded device Nvidia Jetson AGX Xavier. These results show that our method can reduce the energy consumption of depth estimation DNNs while maintaining their accuracy.
- Open Access
- 27 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 1-8
Dynamic Surface Capture for Human Performance by Fusion of Silhouette and Multi-view Stereo
We present a multi-camera based 3D dynamic surface capture solution, which supports high-fidelity generation of 3D dynamic content for various performance scenes. Our system uses a set of RGB cameras to synchronously acquire scene image sequences and performs a processing pipeline to produce 4D videos. We propose a multi-view point cloud reconstruction method which integrates volumetric based guidence and contraint into a coarse-to-fine depth estimation framework. It gives accurate point cloud models and can handle well with the scenes where textureless objects and multiple subjects present. We also present new methods and introduce modifications for several other key computation modules of the processing pipeline, including foreground segmentation, multi-camera calibration, mesh surface reconstruction and registration, texturing, 4D video compression, etc. Experiments on captured scenes show that our system can produce highly accurate and realistic 4D models of the human performances. We develop a 4D video player and toolkit plugins, and demonstrate the use of integrating the 4D content in VR and AR applications.
- 27 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 1-8
Open AccessDynamic Surface Capture for Human Performance by Fusion of Silhouette and Multi-view Stereo
We present a multi-camera based 3D dynamic surface capture solution, which supports high-fidelity generation of 3D dynamic content for various performance scenes. Our system uses a set of RGB cameras to synchronously acquire scene image sequences and performs a processing pipeline to produce 4D videos. We propose a multi-view point cloud reconstruction method which integrates volumetric based guidence and contraint into a coarse-to-fine depth estimation framework. It gives accurate point cloud models and can handle well with the scenes where textureless objects and multiple subjects present. We also present new methods and introduce modifications for several other key computation modules of the processing pipeline, including foreground segmentation, multi-camera calibration, mesh surface reconstruction and registration, texturing, 4D video compression, etc. Experiments on captured scenes show that our system can produce highly accurate and realistic 4D models of the human performances. We develop a 4D video player and toolkit plugins, and demonstrate the use of integrating the 4D content in VR and AR applications.
- 6 October 2022
- Published by Wiley in Medical Physics
- Vol. 50 (3) , 1528-1538
A novel adaptive cubic quasi‐newton optimizer for deep learning based medical image analysis tasks, validated on detection of COVID‐19 and segmentation for COVID‐19 lung infection, liver tumor, and optic disc/cup
Background: Most of existing deep learning research in medical image analysis is focused on networks with stronger performance. These networks have achieved success, while their architectures are complex and even contain massive parameters ranging from thousands to millions in numbers. The nature of high dimension and nonconvex makes it easy to train a suboptimal model through the popular stochastic first-order optimizers, which only use gradient information. Purpose Our purpose is to design an adaptive cubic quasi-Newton optimizer, which could help to escape from suboptimal solution and improve the performance of deep neural networks on four medical image analysis tasks including: detection of COVID-19, COVID-19 lung infection segmentation, liver tumor segmentation, optic disc/cup segmentation. Methods: In this work, we introduce a novel adaptive cubic quasi-Newton optimizer with high-order moment (termed ACQN-H) for medical image analysis. The optimizer dynamically captures the curvature of the loss function by diagonally approximated Hessian and the norm of difference between previous two estimates, which helps to escape from saddle points more efficiently. In addition, to reduce the variance introduced by the stochastic nature of the problem, ACQN-H hires high-order moment through exponential moving average on iteratively calculated approximated Hessian matrix. Extensive experiments are performed to access the performance of ACQN-H. These include detection of COVID-19 using COVID-Net on dataset COVID-chestxray, which contains 16 565 training samples and 1841 test samples; COVID-19 lung infection segmentation using Inf-Net on COVID-CT, which contains 45, 5, and 5 computer tomography (CT) images for training, validation, and testing, respectively; liver tumor segmentation using ResUNet on LiTS2017, which consists of 50 622 abdominal scan images for training and 26 608 images for testing; optic disc/cup segmentation using MRNet on RIGA, which has 655 color fundus images for training and 95 for testing. The results are compared with commonly used stochastic first-order optimizers such as Adam, SGD, and AdaBound, and recently proposed stochastic quasi-Newton optimizer Apollo. In task detection of COVID-19, we use classification accuracy as the evaluation metric. For the other three medical image segmentation tasks, seven commonly used evaluation metrics are utilized, that is, Dice, structure measure, enhanced-alignment measure (EM), mean absolute error (MAE), intersection over union (IoU), true positive rate (TPR), and true negative rate. Results: Experiments on four tasks show that ACQN-H achieves improvements over other stochastic optimizers: (1) comparing with AdaBound, ACQN-H achieves 0.49%, 0.11%, and 0.70% higher accuracy on the COVID-chestxray dataset using network COVID-Net with VGG16, ResNet50 and DenseNet121 as backbones, respectively; (2) ACQN-H has the best scores in terms of evaluation metrics Dice, TPR, EM, and MAE on COVID-CT dataset using network Inf-Net. Particularly, ACQN-H achieves 1.0% better Dice as compared to Apollo; (3) ACQN-H achieves the best results on LiTS2017 dataset using network ResUNet, and outperforms Adam in terms of Dice by 2.3%; (4) ACQN-H improves the performance of network MRNet on RIGA dataset, and achieves 0.5% and 1.0% better scores on cup segmentation for Dice and IoU, respectively, compared with SGD. We also present fivefold validation results of four tasks. It can be found that the results on detection of COVID-19, liver tumor segmentation and optic disc/cup segmentation can achieve high performance with low variance. For COVID-19 lung infection segmentation, the variance on test set is much larger than on validation set, which may due to small size of dataset. Conclusions: The proposed optimizer ACQN-H has been validated on four medical image analysis tasks including: detection of COVID-19 using COVID-Net on COVID-chestxray, COVID-19 lung infection segmentation using Inf-Net on COVID-CT, liver tumor segmentation using ResUNet on LiTS2017, optic disc/cup segmentation using MRNet on RIGA. Experiments show that ACQN-H can achieve some performance improvement. Moreover, the work is expected to boost the performance of existing deep learning networks in medical image analysis.
- 6 October 2022
- Published by Wiley in Medical Physics
- Vol. 50 (3) , 1528-1538
A novel adaptive cubic quasi‐newton optimizer for deep learning based medical image analysis tasks, validated on detection of COVID‐19 and segmentation for COVID‐19 lung infection, liver tumor, and optic disc/cup
Background: Most of existing deep learning research in medical image analysis is focused on networks with stronger performance. These networks have achieved success, while their architectures are complex and even contain massive parameters ranging from thousands to millions in numbers. The nature of high dimension and nonconvex makes it easy to train a suboptimal model through the popular stochastic first-order optimizers, which only use gradient information. Purpose Our purpose is to design an adaptive cubic quasi-Newton optimizer, which could help to escape from suboptimal solution and improve the performance of deep neural networks on four medical image analysis tasks including: detection of COVID-19, COVID-19 lung infection segmentation, liver tumor segmentation, optic disc/cup segmentation. Methods: In this work, we introduce a novel adaptive cubic quasi-Newton optimizer with high-order moment (termed ACQN-H) for medical image analysis. The optimizer dynamically captures the curvature of the loss function by diagonally approximated Hessian and the norm of difference between previous two estimates, which helps to escape from saddle points more efficiently. In addition, to reduce the variance introduced by the stochastic nature of the problem, ACQN-H hires high-order moment through exponential moving average on iteratively calculated approximated Hessian matrix. Extensive experiments are performed to access the performance of ACQN-H. These include detection of COVID-19 using COVID-Net on dataset COVID-chestxray, which contains 16 565 training samples and 1841 test samples; COVID-19 lung infection segmentation using Inf-Net on COVID-CT, which contains 45, 5, and 5 computer tomography (CT) images for training, validation, and testing, respectively; liver tumor segmentation using ResUNet on LiTS2017, which consists of 50 622 abdominal scan images for training and 26 608 images for testing; optic disc/cup segmentation using MRNet on RIGA, which has 655 color fundus images for training and 95 for testing. The results are compared with commonly used stochastic first-order optimizers such as Adam, SGD, and AdaBound, and recently proposed stochastic quasi-Newton optimizer Apollo. In task detection of COVID-19, we use classification accuracy as the evaluation metric. For the other three medical image segmentation tasks, seven commonly used evaluation metrics are utilized, that is, Dice, structure measure, enhanced-alignment measure (EM), mean absolute error (MAE), intersection over union (IoU), true positive rate (TPR), and true negative rate. Results: Experiments on four tasks show that ACQN-H achieves improvements over other stochastic optimizers: (1) comparing with AdaBound, ACQN-H achieves 0.49%, 0.11%, and 0.70% higher accuracy on the COVID-chestxray dataset using network COVID-Net with VGG16, ResNet50 and DenseNet121 as backbones, respectively; (2) ACQN-H has the best scores in terms of evaluation metrics Dice, TPR, EM, and MAE on COVID-CT dataset using network Inf-Net. Particularly, ACQN-H achieves 1.0% better Dice as compared to Apollo; (3) ACQN-H achieves the best results on LiTS2017 dataset using network ResUNet, and outperforms Adam in terms of Dice by 2.3%; (4) ACQN-H improves the performance of network MRNet on RIGA dataset, and achieves 0.5% and 1.0% better scores on cup segmentation for Dice and IoU, respectively, compared with SGD. We also present fivefold validation results of four tasks. It can be found that the results on detection of COVID-19, liver tumor segmentation and optic disc/cup segmentation can achieve high performance with low variance. For COVID-19 lung infection segmentation, the variance on test set is much larger than on validation set, which may due to small size of dataset. Conclusions: The proposed optimizer ACQN-H has been validated on four medical image analysis tasks including: detection of COVID-19 using COVID-Net on COVID-chestxray, COVID-19 lung infection segmentation using Inf-Net on COVID-CT, liver tumor segmentation using ResUNet on LiTS2017, optic disc/cup segmentation using MRNet on RIGA. Experiments show that ACQN-H can achieve some performance improvement. Moreover, the work is expected to boost the performance of existing deep learning networks in medical image analysis.
- Open Access
- 1 January 2021
- Published by Wiley in Computational Intelligence and Neuroscience
- Vol. 2021 (1)
AdaCN: An Adaptive Cubic Newton Method for Nonconvex Stochastic Optimization
In this work, we introduce AdaCN, a novel adaptive cubic Newton method for nonconvex stochastic optimization. AdaCN dynamically captures the curvature of the loss landscape by diagonally approximated Hessian plus the norm of difference between previous two estimates. It only requires at most first order gradients and updates with linear complexity for both time and memory. In order to reduce the variance introduced by the stochastic nature of the problem, AdaCN hires the first and second moment to implement and exponential moving average on iteratively updated stochastic gradients and approximated stochastic Hessians, respectively. We validate AdaCN in extensive experiments, showing that it outperforms other stochastic first order methods (including SGD, Adam, and AdaBound) and stochastic quasi‐Newton method (i.e., Apollo), in terms of both convergence speed and generalization performance.
- 1 January 2021
- Published by Wiley in Computational Intelligence and Neuroscience
- Vol. 2021 (1)
Open AccessAdaCN: An Adaptive Cubic Newton Method for Nonconvex Stochastic Optimization
In this work, we introduce AdaCN, a novel adaptive cubic Newton method for nonconvex stochastic optimization. AdaCN dynamically captures the curvature of the loss landscape by diagonally approximated Hessian plus the norm of difference between previous two estimates. It only requires at most first order gradients and updates with linear complexity for both time and memory. In order to reduce the variance introduced by the stochastic nature of the problem, AdaCN hires the first and second moment to implement and exponential moving average on iteratively updated stochastic gradients and approximated stochastic Hessians, respectively. We validate AdaCN in extensive experiments, showing that it outperforms other stochastic first order methods (including SGD, Adam, and AdaBound) and stochastic quasi‐Newton method (i.e., Apollo), in terms of both convergence speed and generalization performance.