Yuan Wang
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
Yuan Wang has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Ming Liu
Deakin University, Melbourne, VIC, Australia
4 shared publications - Rui Fan
College of Electronics and Information Engineering, Tongji University, Shanghai, China
3 shared publications - Hengli Wang
Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Hong Kong, SAR, China
2 shared publications Ioannis Pitas
School of Informatics, University of Thessaloniki, 541 24 Thessaloniki, Greece
2 shared publications- Bohuan Xue
South China Normal University
1 shared publication - Brian P. Hobbs
Dell Medical School The University of Texas at Austin Austin Texas USA
1 shared publication
Latest Publications
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Data Association Method Based on Pose Refinement and Mismatch Rejection for HD map Localization
Robust and high-precision real-time localization is one of the fundamental requirements for autonomous vehicles. Traditional localization methods based on sensor fusion, vision and LiDAR, suffer from drawbacks like odometry drift and overly dense point cloud maps, which limit their widespread application in autonomous driving tasks. However, in current localization methods depending on HD map, data association faces challenges. These challenges include issues of statistical consistency between features and HD map, as well as mismatches between individual entities of them, which result from large searching pose errors and the occurrence of repeated landmarks. This paper proposes a robust data association method within a semantic localization framework based on HD map. Firstly, the paper refines the coarse pose of the vehicle using the Cruising Model, according to the localization results from the previous two frames, and addresses the statistical consistency issue between semantic features and HD map. Additionally, a RANSAC-based mismatch elimination method is introduced to resolve feature mismatches. Finally, the proposed algorithm is validated on the Waymo dataset. Experimental results demonstrate a matching accuracy of 98.36%, with Euler angle and position errors both below 0.02 degrees and 2 cm, respectively.
- 27 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Data Association Method Based on Pose Refinement and Mismatch Rejection for HD map Localization
Robust and high-precision real-time localization is one of the fundamental requirements for autonomous vehicles. Traditional localization methods based on sensor fusion, vision and LiDAR, suffer from drawbacks like odometry drift and overly dense point cloud maps, which limit their widespread application in autonomous driving tasks. However, in current localization methods depending on HD map, data association faces challenges. These challenges include issues of statistical consistency between features and HD map, as well as mismatches between individual entities of them, which result from large searching pose errors and the occurrence of repeated landmarks. This paper proposes a robust data association method within a semantic localization framework based on HD map. Firstly, the paper refines the coarse pose of the vehicle using the Cruising Model, according to the localization results from the previous two frames, and addresses the statistical consistency issue between semantic features and HD map. Additionally, a RANSAC-based mismatch elimination method is introduced to resolve feature mismatches. Finally, the proposed algorithm is validated on the Waymo dataset. Experimental results demonstrate a matching accuracy of 98.36%, with Euler angle and position errors both below 0.02 degrees and 2 cm, respectively.
- 15 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Analysis Method of Transformer Quality Event Feature Label Based on BERT
Transformer is an indispensable main electrical equipment in power transmission and distribution system. A large number of typical quality event analysis reports are accumulated in daily operation and maintenance, most of which are unstructured descriptive text data. The records have distinct personal characteristics, and it is difficult to understand or analyze by means of automation. For solving the above problem, this paper proposed an analysis method for labeling transformer quality event features based on Bidirectional Encoder Representation from Transformers(BERT). Firstly, the transformer quality event manually extracted feature label is used as input for text vectorization, and the model parameters are obtained by repeatedly using the text get into BERT pre-training model. Then, a training method that changes each learning rate by gradual attenuation is proposed to optimize the model. K-means++ clustering algorithm was used to analysis and Figure out all kinds of quality events. Finally, an example analysis based on 452 actual transformer quality event analysis reports was conducted. The example results showed that the method proposed in this paper achieved a good correspondence between the feature labels of transformer quality events, and the accuracy reached 90%, which is which is at least 4.2% higher than that of the traditional machine learning text data mining technology.
- 15 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Analysis Method of Transformer Quality Event Feature Label Based on BERT
Transformer is an indispensable main electrical equipment in power transmission and distribution system. A large number of typical quality event analysis reports are accumulated in daily operation and maintenance, most of which are unstructured descriptive text data. The records have distinct personal characteristics, and it is difficult to understand or analyze by means of automation. For solving the above problem, this paper proposed an analysis method for labeling transformer quality event features based on Bidirectional Encoder Representation from Transformers(BERT). Firstly, the transformer quality event manually extracted feature label is used as input for text vectorization, and the model parameters are obtained by repeatedly using the text get into BERT pre-training model. Then, a training method that changes each learning rate by gradual attenuation is proposed to optimize the model. K-means++ clustering algorithm was used to analysis and Figure out all kinds of quality events. Finally, an example analysis based on 452 actual transformer quality event analysis reports was conducted. The example results showed that the method proposed in this paper achieved a good correspondence between the feature labels of transformer quality events, and the accuracy reached 90%, which is which is at least 4.2% higher than that of the traditional machine learning text data mining technology.
- 24 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 8144-8154
Graph Attention Layer Evolves Semantic Segmentation for Road Pothole Detection: A Benchmark and Algorithms
Existing road pothole detection approaches can be classified as computer vision-based or machine learning-based. The former approaches typically employ 2-D image analysis/understanding or 3-D point cloud modeling and segmentation algorithms to detect (i.e., recognize and localize) road potholes from vision sensor data, e.g., RGB images and/or depth/disparity images. The latter approaches generally address road pothole detection using convolutional neural networks (CNNs) in an end-to-end manner. However, road potholes are not necessarily ubiquitous and it is challenging to prepare a large well-annotated dataset for CNN training. In this regard, while computer vision-based methods were the mainstream research trend in the past decade, machine learning-based methods were merely discussed. Recently, we published the first stereo vision-based road pothole detection dataset and a novel disparity transformation algorithm, whereby the damaged and undamaged road areas can be highly distinguished. However, there are no benchmarks currently available for state-of-the-art (SoTA) CNNs trained using either disparity images or transformed disparity images. Therefore, in this paper, we first discuss the SoTA CNNs designed for semantic segmentation and evaluate their performance for road pothole detection with extensive experiments. Additionally, inspired by graph neural network (GNN), we propose a novel CNN layer, referred to as graph attention layer (GAL), which can be easily deployed in any existing CNN to optimize image feature representations for semantic segmentation. Our experiments compare GAL-DeepLabv3+, our best-performing implementation, with nine SoTA CNNs on three modalities of training data: RGB images, disparity images, and transformed disparity images. The experimental results suggest that our proposed GAL-DeepLabv3+ achieves the best overall pothole detection accuracy on all training data modalities. The source code, dataset, and benchmark are publicly available at mias.group/GAL-Pothole-Detection.
- 24 September 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 30 (10577149) , 8144-8154
Graph Attention Layer Evolves Semantic Segmentation for Road Pothole Detection: A Benchmark and Algorithms
Existing road pothole detection approaches can be classified as computer vision-based or machine learning-based. The former approaches typically employ 2-D image analysis/understanding or 3-D point cloud modeling and segmentation algorithms to detect (i.e., recognize and localize) road potholes from vision sensor data, e.g., RGB images and/or depth/disparity images. The latter approaches generally address road pothole detection using convolutional neural networks (CNNs) in an end-to-end manner. However, road potholes are not necessarily ubiquitous and it is challenging to prepare a large well-annotated dataset for CNN training. In this regard, while computer vision-based methods were the mainstream research trend in the past decade, machine learning-based methods were merely discussed. Recently, we published the first stereo vision-based road pothole detection dataset and a novel disparity transformation algorithm, whereby the damaged and undamaged road areas can be highly distinguished. However, there are no benchmarks currently available for state-of-the-art (SoTA) CNNs trained using either disparity images or transformed disparity images. Therefore, in this paper, we first discuss the SoTA CNNs designed for semantic segmentation and evaluate their performance for road pothole detection with extensive experiments. Additionally, inspired by graph neural network (GNN), we propose a novel CNN layer, referred to as graph attention layer (GAL), which can be easily deployed in any existing CNN to optimize image feature representations for semantic segmentation. Our experiments compare GAL-DeepLabv3+, our best-performing implementation, with nine SoTA CNNs on three modalities of training data: RGB images, disparity images, and transformed disparity images. The experimental results suggest that our proposed GAL-DeepLabv3+ achieves the best overall pothole detection accuracy on all training data modalities. The source code, dataset, and benchmark are publicly available at mias.group/GAL-Pothole-Detection.
- 24 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cybernetics
- Vol. 52 (7) , 5799-5808
Rethinking Road Surface 3-D Reconstruction and Pothole Detection: From Perspective Transformation to Disparity Map Segmentation
Potholes are one of the most common forms of road damage, which can severely affect driving comfort, road safety, and vehicle condition. Pothole detection is typically performed by either structural engineers or certified inspectors. However, this task is not only hazardous for the personnel but also extremely time consuming. This article presents an efficient pothole detection algorithm based on road disparity map estimation and segmentation. We first incorporate the stereo rig roll angle into shifting distance calculation to generalize perspective transformation. The road disparities are then efficiently estimated using semiglobal matching. A disparity map transformation algorithm is then performed to better distinguish the damaged road areas. Subsequently, we utilize simple linear iterative clustering to group the transformed disparities into a collection of superpixels. The potholes are finally detected by finding the superpixels, whose intensities are lower than an adaptively determined threshold. The proposed algorithm is implemented on an NVIDIA RTX 2080 Ti GPU in CUDA. The experimental results demonstrate that our proposed road pothole detection algorithm achieves state-of-the-art accuracy and efficiency.
- 24 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cybernetics
- Vol. 52 (7) , 5799-5808
Rethinking Road Surface 3-D Reconstruction and Pothole Detection: From Perspective Transformation to Disparity Map Segmentation
Potholes are one of the most common forms of road damage, which can severely affect driving comfort, road safety, and vehicle condition. Pothole detection is typically performed by either structural engineers or certified inspectors. However, this task is not only hazardous for the personnel but also extremely time consuming. This article presents an efficient pothole detection algorithm based on road disparity map estimation and segmentation. We first incorporate the stereo rig roll angle into shifting distance calculation to generalize perspective transformation. The road disparities are then efficiently estimated using semiglobal matching. A disparity map transformation algorithm is then performed to better distinguish the damaged road areas. Subsequently, we utilize simple linear iterative clustering to group the transformed disparities into a collection of superpixels. The potholes are finally detected by finding the superpixels, whose intensities are lower than an adaptively determined threshold. The proposed algorithm is implemented on an NVIDIA RTX 2080 Ti GPU in CUDA. The experimental results demonstrate that our proposed road pothole detection algorithm achieves state-of-the-art accuracy and efficiency.
- 18 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 6 (3) , 5405-5412
Three-Filters-to-Normal: An Accurate and Ultrafast Surface Normal Estimator
This paper proposes three-filters-to-normal (3F2N), an accurate and ultrafast surface normal estimator (SNE), which is designed for structured range sensor data, e.g., depth/disparity images. 3F2N SNE computes surface normals by simply performing three filtering operations (two image gradient filters in horizontal and vertical directions, respectively, and a mean/median filter) on an inverse depth image or a disparity image (due to that disparity is in inverse proportion to depth). Despite the simplicity of 3F2N SNE, no similar method already exists in the literature. To evaluate the performance of our proposed SNE, we created three large-scale synthetic datasets (easy, medium and hard) using 24 3D mesh models, each of which is used to generate 1800--2500 pairs of depth images (resolution: 480x640 pixels) and the corresponding surface normal ground truth from different views. 3F2N SNE demonstrates the state-of-the-art performance, outperforming all other existing geometry-based SNEs, where the average angular errors with respect to the easy, medium and hard datasets are 1.6 degrees, 5.6 degrees and 15.3 degrees, respectively. Furthermore, our C++ and CUDA implementations achieve a processing speed of over 260 Hz and 21 kHz, respectively. We have also contributed our source code to the OpenCV repository. Our datasets and source code are publicly available at: sites.google.com/view/3f2n.
- 18 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 6 (3) , 5405-5412
Three-Filters-to-Normal: An Accurate and Ultrafast Surface Normal Estimator
This paper proposes three-filters-to-normal (3F2N), an accurate and ultrafast surface normal estimator (SNE), which is designed for structured range sensor data, e.g., depth/disparity images. 3F2N SNE computes surface normals by simply performing three filtering operations (two image gradient filters in horizontal and vertical directions, respectively, and a mean/median filter) on an inverse depth image or a disparity image (due to that disparity is in inverse proportion to depth). Despite the simplicity of 3F2N SNE, no similar method already exists in the literature. To evaluate the performance of our proposed SNE, we created three large-scale synthetic datasets (easy, medium and hard) using 24 3D mesh models, each of which is used to generate 1800--2500 pairs of depth images (resolution: 480x640 pixels) and the corresponding surface normal ground truth from different views. 3F2N SNE demonstrates the state-of-the-art performance, outperforming all other existing geometry-based SNEs, where the average angular errors with respect to the easy, medium and hard datasets are 1.6 degrees, 5.6 degrees and 15.3 degrees, respectively. Furthermore, our C++ and CUDA implementations achieve a processing speed of over 260 Hz and 21 kHz, respectively. We have also contributed our source code to the OpenCV repository. Our datasets and source code are publicly available at: sites.google.com/view/3f2n.
- 23 January 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 4 (2) , 1263-1270
Focal Loss in 3D Object Detection
3D object detection is still an open problem in autonomous driving scenes. When recognizing and localizing key objects from sparse 3D inputs, autonomous vehicles suffer from a larger continuous searching space and higher fore-background imbalance compared to image-based object detection. In this letter, we aim to solve this fore-background imbalance in 3D object detection. Inspired by the recent use of focal loss in image-based object detection, we extend this hard-mining improvement of binary cross entropy to point-cloud-based object detection and conduct experiments to show its performance based on two different 3D detectors: 3D-FCN and VoxelNet. The evaluation results show up to 11.2AP gains through the focal loss in a wide range of hyperparameters for 3D object detection.
- 23 January 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 4 (2) , 1263-1270
Focal Loss in 3D Object Detection
3D object detection is still an open problem in autonomous driving scenes. When recognizing and localizing key objects from sparse 3D inputs, autonomous vehicles suffer from a larger continuous searching space and higher fore-background imbalance compared to image-based object detection. In this letter, we aim to solve this fore-background imbalance in 3D object detection. Inspired by the recent use of focal loss in image-based object detection, we extend this hard-mining improvement of binary cross entropy to point-cloud-based object detection and conduct experiments to show its performance based on two different 3D detectors: 3D-FCN and VoxelNet. The evaluation results show up to 11.2AP gains through the focal loss in a wide range of hyperparameters for 3D object detection.
- Open Access
A Time Series Graph Cut Image Segmentation Scheme for Liver Tumors
Tumor detection in biomedical imaging is a time-consuming process for medical professionals and is not without errors. Thus in recent decades, researchers have developed algorithmic techniques for image processing using a wide variety of mathematical methods, such as statistical modeling, variational techniques, and machine learning. In this paper, we propose a semi-automatic method for liver segmentation of 2D CT scans into three labels denoting healthy, vessel, or tumor tissue based on graph cuts. First, we create a feature vector for each pixel in a novel way that consists of the 59 intensity values in the time series data and propose a simplified perimeter cost term in the energy functional. We normalize the data and perimeter terms in the functional to expedite the graph cut without having to optimize the scaling parameter $\lambda$. In place of a training process, predetermined tissue means are computed based on sample regions identified by expert radiologists. The proposed method also has the advantage of being relatively simple to implement computationally. It was evaluated against the ground truth on a clinical CT dataset of 10 tumors and yielded segmentations with a mean Dice similarity coefficient (DSC) of .77 and mean volume overlap error (VOE) of 36.7%. The average processing time was 1.25 minutes per slice.Open AccessA Time Series Graph Cut Image Segmentation Scheme for Liver Tumors
Tumor detection in biomedical imaging is a time-consuming process for medical professionals and is not without errors. Thus in recent decades, researchers have developed algorithmic techniques for image processing using a wide variety of mathematical methods, such as statistical modeling, variational techniques, and machine learning. In this paper, we propose a semi-automatic method for liver segmentation of 2D CT scans into three labels denoting healthy, vessel, or tumor tissue based on graph cuts. First, we create a feature vector for each pixel in a novel way that consists of the 59 intensity values in the time series data and propose a simplified perimeter cost term in the energy functional. We normalize the data and perimeter terms in the functional to expedite the graph cut without having to optimize the scaling parameter $\lambda$. In place of a training process, predetermined tissue means are computed based on sample regions identified by expert radiologists. The proposed method also has the advantage of being relatively simple to implement computationally. It was evaluated against the ground truth on a clinical CT dataset of 10 tumors and yielded segmentations with a mean Dice similarity coefficient (DSC) of .77 and mean volume overlap error (VOE) of 36.7%. The average processing time was 1.25 minutes per slice.