Yuxing Peng
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
Yuxing Peng has a total of 80 co-authors. Here are the top co-authors by number of shared publications:
- Zhen Huang
PDL Lab, College of Computer Science and Technology, National University of Defense Technology, Changsha, Hunan, China
6 shared publications - Dongsheng Li
PDL Lab, College of Computer Science and Technology, National University of Defense Technology, Changsha, Hunan, China
4 shared publications - Xinwang Liu
National University of Defense Technology
4 shared publications Changjian Wang
National university of defense technology, Changsha, China
3 shared publications- Mingxing Tang
Science and Technology on Parallel and Distributed Laboratory, National University of Defense Technology, Changsha 410000, China
3 shared publications - Yiming Zhang
Shanghai Jiao Tong University, Shanghai, China
3 shared publications
Latest Publications
- Open Access
- 14 August 2023
- Published by Springer Nature in Journal of Cardiothoracic Surgery
- Vol. 18 (1) , 1-8
Endovascular treatment for massive haemoptysis due to pulmonary pseudoaneurysm: report of 23 cases
Purpose: To evaluate the safety and effectiveness of endovascular treatment for massive haemoptysis caused by pulmonary pseudoaneurysm (PAP). Methods: The clinical data, imaging data, and endovascular treatment protocol of 23 patients with massive haemoptysis caused by continuous PAP were retrospectively analysed. The success, complications, postoperative recurrence rate, and influence of the treatment on pulmonary artery pressure were also evaluated. Results: Nineteen patients with a bronchial artery-pulmonary artery (BA-PA) and/or nonbronchial systemic artery-pulmonary artery (NBSA-PA) fistula underwent bronchial artery embolization (BAE) and/or nonbronchial systemic artery embolization (NBSAE) + pulmonary artery embolization (PAE). The pulmonary artery (PA) pressures before and after embolization were 52.11 ± 2.12 (35–69 cmH2O) and 33.58 ± 1.63 (22–44 cmH2O), respectively (P = 0.001). Four patients did not have a BA-PA and/or NBSA-PA fistula. Embolization was performed in two patients with a distal PAP of the pulmonalis lobar arteria. Bare stent-assisted microcoils embolization was performed in the other two patients with a PAP of the main pulmonary lobar arteries. The PA pressures of the four patients before and after treatment were 24.50 ± 1.32 (22–28 cmH2O) and 24.75 ± 1.70 (22–29 cmH2O), respectively (P = 0.850). The technique had a 100% success rate with no serious complications and a postoperative recurrence rate of 30%. Conclusion: Endovascular treatment is safe and effective for massive haemoptysis caused by PAP. BAE and/or NBSAE can effectively reduce pulmonary hypertension in patients with a BA-PA and/or NBSA-PA fistula.
- 14 August 2023
- Published by Springer Nature in Journal of Cardiothoracic Surgery
- Vol. 18 (1) , 1-8
Open AccessEndovascular treatment for massive haemoptysis due to pulmonary pseudoaneurysm: report of 23 cases
Purpose: To evaluate the safety and effectiveness of endovascular treatment for massive haemoptysis caused by pulmonary pseudoaneurysm (PAP). Methods: The clinical data, imaging data, and endovascular treatment protocol of 23 patients with massive haemoptysis caused by continuous PAP were retrospectively analysed. The success, complications, postoperative recurrence rate, and influence of the treatment on pulmonary artery pressure were also evaluated. Results: Nineteen patients with a bronchial artery-pulmonary artery (BA-PA) and/or nonbronchial systemic artery-pulmonary artery (NBSA-PA) fistula underwent bronchial artery embolization (BAE) and/or nonbronchial systemic artery embolization (NBSAE) + pulmonary artery embolization (PAE). The pulmonary artery (PA) pressures before and after embolization were 52.11 ± 2.12 (35–69 cmH2O) and 33.58 ± 1.63 (22–44 cmH2O), respectively (P = 0.001). Four patients did not have a BA-PA and/or NBSA-PA fistula. Embolization was performed in two patients with a distal PAP of the pulmonalis lobar arteria. Bare stent-assisted microcoils embolization was performed in the other two patients with a PAP of the main pulmonary lobar arteries. The PA pressures of the four patients before and after treatment were 24.50 ± 1.32 (22–28 cmH2O) and 24.75 ± 1.70 (22–29 cmH2O), respectively (P = 0.850). The technique had a 100% success rate with no serious complications and a postoperative recurrence rate of 30%. Conclusion: Endovascular treatment is safe and effective for massive haemoptysis caused by PAP. BAE and/or NBSAE can effectively reduce pulmonary hypertension in patients with a BA-PA and/or NBSA-PA fistula.
- Top Cited Papers
- 20 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 45 (8) , 9806-9821
GeoTransformer: Fast and Robust Point Cloud Registration With Geometric Transformer
We study the problem of extracting accurate correspondences for point cloud registration. Recent keypoint-free methods have shown great potential through bypassing the detection of repeatable keypoints which is difficult to do especially in low-overlap scenarios. They seek correspondences over downsampled superpoints, which are then propagated to dense points. Superpoints are matched based on whether their neighboring patches overlap. Such sparse and loose matching requires contextual features capturing the geometric structure of the point clouds. We propose Geometric Transformer, or GeoTransformer for short, to learn geometric feature for robust superpoint matching. It encodes pair-wise distances and triplet-wise angles, making it invariant to rigid transformation and robust in low-overlap cases. The simplistic design attains surprisingly high matching accuracy such that no RANSAC is required in the estimation of alignment transformation, leading to 100 times acceleration. Extensive experiments on rich benchmarks encompassing indoor, outdoor, synthetic, multiway and non-rigid demonstrate the efficacy of GeoTransformer. Notably, our method improves the inlier ratio by$18{\sim }31$ percentage points and the registration recall by over 7 points on the challenging 3DLoMatch benchmark.- 20 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 45 (8) , 9806-9821
Top Cited PapersGeoTransformer: Fast and Robust Point Cloud Registration With Geometric Transformer
We study the problem of extracting accurate correspondences for point cloud registration. Recent keypoint-free methods have shown great potential through bypassing the detection of repeatable keypoints which is difficult to do especially in low-overlap scenarios. They seek correspondences over downsampled superpoints, which are then propagated to dense points. Superpoints are matched based on whether their neighboring patches overlap. Such sparse and loose matching requires contextual features capturing the geometric structure of the point clouds. We propose Geometric Transformer, or GeoTransformer for short, to learn geometric feature for robust superpoint matching. It encodes pair-wise distances and triplet-wise angles, making it invariant to rigid transformation and robust in low-overlap cases. The simplistic design attains surprisingly high matching accuracy such that no RANSAC is required in the estimation of alignment transformation, leading to 100 times acceleration. Extensive experiments on rich benchmarks encompassing indoor, outdoor, synthetic, multiway and non-rigid demonstrate the efficacy of GeoTransformer. Notably, our method improves the inlier ratio by$18{\sim }31$ percentage points and the registration recall by over 7 points on the challenging 3DLoMatch benchmark. - Open Access
- 1 February 2021
- Published by Association for Computing Machinery (ACM) in Proceedings of the VLDB Endowment
- Vol. 14 (6) , 864-877
ParaX
Despite the fact that GPUs and accelerators are more efficient in deep learning (DL), commercial clouds like Facebook and Amazon now heavily use CPUs in DL computation because there are large numbers of CPUs which would otherwise sit idle during off-peak periods. Following the trend, CPU vendors have not only released high-performance many-core CPUs but also developed efficient math kernel libraries. However, current DL platforms cannot scale well to a large number of CPU cores, making many-core CPUs inefficient in DL computation. We analyze the memory access patterns of various layers and identify the root cause of the low scalability, i.e., the per-layer barriers that are implicitly imposed by current platforms which assign one single instance (i.e., one batch of input data) to a CPU. The barriers cause severe memory bandwidth contention and CPU starvation in the access-intensive layers (like activation and BN). This paper presents a novel approach called ParaX, which boosts the performance of DL on many-core CPUs by effectively alleviating bandwidth contention and CPU starvation. Our key idea is to assign one instance to each CPU core instead of to the entire CPU, so as to remove the per-layer barriers on the executions of the many cores. ParaX designs an ultralight scheduling policy which sufficiently overlaps the access-intensive layers with the compute-intensive ones to avoid contention, and proposes a NUMA-aware gradient server mechanism for training which leverages shared memory to substantially reduce the overhead of per-iteration parameter synchronization. We have implemented ParaX on MXNet. Extensive evaluation on a two-NUMA Intel 8280 CPU shows that ParaX significantly improves the training/inference throughput for all tested models (for image recognition and natural language processing) by 1.73X ~ 2.93X.
- 1 February 2021
- Published by Association for Computing Machinery (ACM) in Proceedings of the VLDB Endowment
- Vol. 14 (6) , 864-877
Open AccessParaX
Despite the fact that GPUs and accelerators are more efficient in deep learning (DL), commercial clouds like Facebook and Amazon now heavily use CPUs in DL computation because there are large numbers of CPUs which would otherwise sit idle during off-peak periods. Following the trend, CPU vendors have not only released high-performance many-core CPUs but also developed efficient math kernel libraries. However, current DL platforms cannot scale well to a large number of CPU cores, making many-core CPUs inefficient in DL computation. We analyze the memory access patterns of various layers and identify the root cause of the low scalability, i.e., the per-layer barriers that are implicitly imposed by current platforms which assign one single instance (i.e., one batch of input data) to a CPU. The barriers cause severe memory bandwidth contention and CPU starvation in the access-intensive layers (like activation and BN). This paper presents a novel approach called ParaX, which boosts the performance of DL on many-core CPUs by effectively alleviating bandwidth contention and CPU starvation. Our key idea is to assign one instance to each CPU core instead of to the entire CPU, so as to remove the per-layer barriers on the executions of the many cores. ParaX designs an ultralight scheduling policy which sufficiently overlaps the access-intensive layers with the compute-intensive ones to avoid contention, and proposes a NUMA-aware gradient server mechanism for training which leverages shared memory to substantially reduce the overhead of per-iteration parameter synchronization. We have implemented ParaX on MXNet. Extensive evaluation on a two-NUMA Intel 8280 CPU shows that ParaX significantly improves the training/inference throughput for all tested models (for image recognition and natural language processing) by 1.73X ~ 2.93X.
- 1 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in China Communications
- Vol. 17 (9) , 77-89
A distributed computing framework based on lightweight variance reduction method to accelerate machine learning training on blockchain
- 1 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in China Communications
- Vol. 17 (9) , 77-89
A distributed computing framework based on lightweight variance reduction method to accelerate machine learning training on blockchain
- Open Access
- 1 September 2019
- Published by MDPI AG in Applied Sciences
- Vol. 9 (17) , 3569
A Bounded Scheduling Method for Adaptive Gradient Methods
Many adaptive gradient methods have been successfully applied to train deep neural networks, such as Adagrad, Adadelta, RMSprop and Adam. These methods perform local optimization with an element-wise scaling learning rate based on past gradients. Although these methods can achieve an advantageous training loss, some researchers have pointed out that their generalization capability tends to be poor as compared to stochastic gradient descent (SGD) in many applications. These methods obtain a rapid initial training process but fail to converge to an optimal solution due to the unstable and extreme learning rates. In this paper, we investigate the adaptive gradient methods and get the insights on various factors that may lead to poor performance of Adam. To overcome that, we propose a bounded scheduling algorithm for Adam, which can not only improve the generalization capability but also ensure the convergence. To validate our claims, we carry out a series of experiments on the image classification and the language modeling tasks on several standard benchmarks such as ResNet, DenseNet, SENet and LSTM on typical data sets such as CIFAR-10, CIFAR-100 and Penn Treebank. Experimental results show that our method can eliminate the generalization gap between Adam and SGD, meanwhile maintaining a relative high convergence rate during training.
- 1 September 2019
- Published by MDPI AG in Applied Sciences
- Vol. 9 (17) , 3569
Open AccessA Bounded Scheduling Method for Adaptive Gradient Methods
Many adaptive gradient methods have been successfully applied to train deep neural networks, such as Adagrad, Adadelta, RMSprop and Adam. These methods perform local optimization with an element-wise scaling learning rate based on past gradients. Although these methods can achieve an advantageous training loss, some researchers have pointed out that their generalization capability tends to be poor as compared to stochastic gradient descent (SGD) in many applications. These methods obtain a rapid initial training process but fail to converge to an optimal solution due to the unstable and extreme learning rates. In this paper, we investigate the adaptive gradient methods and get the insights on various factors that may lead to poor performance of Adam. To overcome that, we propose a bounded scheduling algorithm for Adam, which can not only improve the generalization capability but also ensure the convergence. To validate our claims, we carry out a series of experiments on the image classification and the language modeling tasks on several standard benchmarks such as ResNet, DenseNet, SENet and LSTM on typical data sets such as CIFAR-10, CIFAR-100 and Penn Treebank. Experimental results show that our method can eliminate the generalization gap between Adam and SGD, meanwhile maintaining a relative high convergence rate during training.
- Open Access
A Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning
Rapid progress has been made in the field of reading comprehension and question answering, where several systems have achieved human parity in some simplified settings. However, the performance of these models degrades significantly when they are applied to more realistic scenarios, such as answers involve various types, multiple text strings are correct answers, or discrete reasoning abilities are required. In this paper, we introduce the Multi-Type Multi-Span Network (MTMSN), a neural reading comprehension model that combines a multi-type answer predictor designed to support various answer types (e.g., span, count, negation, and arithmetic expression) with a multi-span extraction method for dynamically producing one or multiple text spans. In addition, an arithmetic expression reranking mechanism is proposed to rank expression candidates for further confirming the prediction. Experiments show that our model achieves 79.9 F1 on the DROP hidden test set, creating new state-of-the-art results. Source code\footnote{\url{https://github.com/huminghao16/MTMSN}} is released to facilitate future work.
Open AccessA Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning
Rapid progress has been made in the field of reading comprehension and question answering, where several systems have achieved human parity in some simplified settings. However, the performance of these models degrades significantly when they are applied to more realistic scenarios, such as answers involve various types, multiple text strings are correct answers, or discrete reasoning abilities are required. In this paper, we introduce the Multi-Type Multi-Span Network (MTMSN), a neural reading comprehension model that combines a multi-type answer predictor designed to support various answer types (e.g., span, count, negation, and arithmetic expression) with a multi-span extraction method for dynamically producing one or multiple text spans. In addition, an arithmetic expression reranking mechanism is proposed to rank expression candidates for further confirming the prediction. Experiments show that our model achieves 79.9 F1 on the DROP hidden test set, creating new state-of-the-art results. Source code\footnote{\url{https://github.com/huminghao16/MTMSN}} is released to facilitate future work.
- 24 June 2019
- Published by Springer Nature in Neural Computing and Applications
- Vol. 32 (12) , 8089-8100
Accelerating SGD using flexible variance reduction on large-scale datasets
- 24 June 2019
- Published by Springer Nature in Neural Computing and Applications
- Vol. 32 (12) , 8089-8100
Accelerating SGD using flexible variance reduction on large-scale datasets
- Open Access
- 1 January 2019
- Published by Association for Computational Linguistics (ACL)
- p. 1596-1606
A Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning
Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.
- 1 January 2019
- Published by Association for Computational Linguistics (ACL)
- p. 1596-1606
Open AccessA Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning
Minghao Hu, Yuxing Peng, Zhen Huang, Dongsheng Li. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.
- Open Access
A Semi-Supervised Model for Top-N Recommendation
Top-N recommendation is an important recommendation technique that delivers a ranked top-N item list to each user. Data sparsity is a great challenge for top-N recommendation. In order to tackle this problem, in this paper, we propose a semi-supervised model called Semi-BPR (Semi-Supervised Bayesian Personalized Ranking). Our approach is based on the assumption that, for a given model, users always prefer items ranked higher in the generated recommendation list. Therefore, we select a certain number of items ranked higher in the recommendation list to construct an intermediate set and optimize the metric Area Under the Curve (AUC). In addition, we treat the intermediate set as a teaching set and design a semi-supervised self-training model. We conduct a series of experiments on three popular datasets to compare the proposed approach with several state-of-the-art baselines. The experimental results demonstrate that our approach significantly outperforms the other methods for all evaluation metrics, especially for sparse datasets.
Open AccessA Semi-Supervised Model for Top-N Recommendation
Top-N recommendation is an important recommendation technique that delivers a ranked top-N item list to each user. Data sparsity is a great challenge for top-N recommendation. In order to tackle this problem, in this paper, we propose a semi-supervised model called Semi-BPR (Semi-Supervised Bayesian Personalized Ranking). Our approach is based on the assumption that, for a given model, users always prefer items ranked higher in the generated recommendation list. Therefore, we select a certain number of items ranked higher in the recommendation list to construct an intermediate set and optimize the metric Area Under the Curve (AUC). In addition, we treat the intermediate set as a teaching set and design a semi-supervised self-training model. We conduct a series of experiments on three popular datasets to compare the proposed approach with several state-of-the-art baselines. The experimental results demonstrate that our approach significantly outperforms the other methods for all evaluation metrics, especially for sparse datasets.
- Open Access
- 12 September 2018
- Published by Wiley in Computational Intelligence and Neuroscience
- Vol. 2018, 1
Corrigendum to “A Composite Model of Wound Segmentation Based on Traditional Methods and Deep Neural Networks”
[This corrects the article DOI: 10.1155/2018/4149103.].
- 12 September 2018
- Published by Wiley in Computational Intelligence and Neuroscience
- Vol. 2018, 1
Open AccessCorrigendum to “A Composite Model of Wound Segmentation Based on Traditional Methods and Deep Neural Networks”
[This corrects the article DOI: 10.1155/2018/4149103.].