Haoran Yin
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
Haoran Yin has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Atuganile Malambugi
University of Dar es Salaam, Dar es Salaam, Tanzania | Tanzania Fisheries Research Institute, Dar es Salaam Tanzania
1 shared publication - Cong Yang
School of Future Science and Engineering, Soochow University, Suzhou, China
1 shared publication Guoqing Lu
1 shared publication- Jiajun Sun
Center for Research on Intelligent Perception and Computing (CRIPAC), National Laboratory of Pattern Recognition (NLPR), Institute of Automation, Chinese Academy of Sciences (CASIA), Beijing, China
1 shared publication Jiaxin Zhang
BeeLab, School of Future Science and Engineering, Soochow University, Suzhou, China
1 shared publication- Junran Peng
University of Science and Technology Beijing, Beijing, China
1 shared publication
Latest Publications
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (7) , 8158-8172
CAMAv2: A Vision-Centric Approach for Static Map Element Annotation
The recent advancement of Bird’s Eye View (BEV) perception algorithms requires extensive, high-quality annotated map data for effective training and deployment in real-world scenarios. However, existing HD map-based auto-labeling methods, like those found in public datasets, face significant issues related to efficiency and accuracy. For instance, the nuScenes dataset reveals considerable misalignment and inconsistency between images and their annotations, with an average reprojection error of approximately 8.03 pixels. Additionally, the dependence on HD maps limits the applicability of these methods for large-scale, real-world auto-labeling. To tackle these challenges, we introduce CAMAv2: a vision-centric approach for Consistent and Accurate Map Annotation. This pipeline primarily utilizes camera inputs to generate precise 3D annotations of static map elements, achieving high reprojection accuracy across all surrounding cameras and maintaining spatiotemporal consistency throughout the entire sequence. Importantly, CAMAv2 annotations show lower reprojection errors compared to the original nuScenes map elements, with errors of 4.96 pixels versus 8.03 pixels. Comprehensive evaluations across various public datasets confirm the feasibility and generalizability of our pipeline in diverse environments, as well as under different weather and lighting conditions.
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (7) , 8158-8172
CAMAv2: A Vision-Centric Approach for Static Map Element Annotation
The recent advancement of Bird’s Eye View (BEV) perception algorithms requires extensive, high-quality annotated map data for effective training and deployment in real-world scenarios. However, existing HD map-based auto-labeling methods, like those found in public datasets, face significant issues related to efficiency and accuracy. For instance, the nuScenes dataset reveals considerable misalignment and inconsistency between images and their annotations, with an average reprojection error of approximately 8.03 pixels. Additionally, the dependence on HD maps limits the applicability of these methods for large-scale, real-world auto-labeling. To tackle these challenges, we introduce CAMAv2: a vision-centric approach for Consistent and Accurate Map Annotation. This pipeline primarily utilizes camera inputs to generate precise 3D annotations of static map elements, achieving high reprojection accuracy across all surrounding cameras and maintaining spatiotemporal consistency throughout the entire sequence. Importantly, CAMAv2 annotations show lower reprojection errors compared to the original nuScenes map elements, with errors of 4.96 pixels versus 8.03 pixels. Comprehensive evaluations across various public datasets confirm the feasibility and generalizability of our pipeline in diverse environments, as well as under different weather and lighting conditions.
- 16 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 45 (10) , 11856-11868
GAIA-Universe: Everything is Super-Netify
Pre-training on large-scale datasets has played an increasingly significant role in computer vision and natural language processing recently. However, as there exist numerous application scenarios that have distinctive demands such as certain latency constraints and specialized data distributions, it is prohibitively expensive to take advantage of large-scale pre-training for per-task requirements. we focus on two fundamental perception tasks (object detection and semantic segmentation) and present a complete and flexible system named GAIA-Universe(GAIA), which could automatically and efficiently give birth to customized solutions according to heterogeneous downstream needs through data union and super-net training. GAIA is capable of providing powerful pre-trained weights and searching models that conform to downstream demands such as hardware constraints, computation constraints, specified data domains, and telling relevant data for practitioners who have very few datapoints on their tasks. With GAIA, we achieve promising results on COCO, Objects365, Open Images, BDD100k, and UODB which is a collection of datasets including KITTI, VOC, WiderFace, DOTA, Clipart, Comic, and more. Taking COCO as an example, GAIA is able to efficiently produce models covering a wide range of latency from 16ms to 53ms, and yields AP from 38.2 to 46.5 without whistles and bells. GAIA is released at https://github.com/GAIA-vision .
- 16 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 45 (10) , 11856-11868
GAIA-Universe: Everything is Super-Netify
Pre-training on large-scale datasets has played an increasingly significant role in computer vision and natural language processing recently. However, as there exist numerous application scenarios that have distinctive demands such as certain latency constraints and specialized data distributions, it is prohibitively expensive to take advantage of large-scale pre-training for per-task requirements. we focus on two fundamental perception tasks (object detection and semantic segmentation) and present a complete and flexible system named GAIA-Universe(GAIA), which could automatically and efficiently give birth to customized solutions according to heterogeneous downstream needs through data union and super-net training. GAIA is capable of providing powerful pre-trained weights and searching models that conform to downstream demands such as hardware constraints, computation constraints, specified data domains, and telling relevant data for practitioners who have very few datapoints on their tasks. With GAIA, we achieve promising results on COCO, Objects365, Open Images, BDD100k, and UODB which is a collection of datasets including KITTI, VOC, WiderFace, DOTA, Clipart, Comic, and more. Taking COCO as an example, GAIA is able to efficiently produce models covering a wide range of latency from 16ms to 53ms, and yields AP from 38.2 to 46.5 without whistles and bells. GAIA is released at https://github.com/GAIA-vision .
- Open Access
- 18 June 2022
- Published by Wiley in Wireless Communications and Mobile Computing
- Vol. 2022, 1-10
Identification of Aircraft Wake Vortex Based on VGGNet
The exploration of information for aircraft wake vortex enables us to obtain new knowledge of wake turbulence separation standards. Traditional manual methods cannot work satisfactorily for the identification of great number of wake vortex data with high accuracy. Fortunately, the LiDAR intensity data can be explained by integrating LiDAR products with the strategies of computer vision. To overcome the limitation of traditional manual methods, this paper is aimed at developing an automatic method to identify a given set of wake vortices from various aircrafts. The main innovation works are outlined as follows. (1) From the wake vortex data that consists of various aircrafts measured by Wind3D 6000 LiDAR, a grayscale dataset of wake flow is constructed to boost the deep learning model for identifying aircraft wake vortex. (2) Following this, we propose a new method for the identification of aircraft wake vortex by modifying the VGG16 network, providing binary classifications of uncertain behavior patterns for wake vortices. To evaluate the proposed identification model, performance evaluation was conducted on our dataset, where experimental results revealed the values of 0.984, 0.951, 0.959, and 0.955 in terms of accuracy, precision, recall, and F1-score, respectively.
- 18 June 2022
- Published by Wiley in Wireless Communications and Mobile Computing
- Vol. 2022, 1-10
Open AccessIdentification of Aircraft Wake Vortex Based on VGGNet
The exploration of information for aircraft wake vortex enables us to obtain new knowledge of wake turbulence separation standards. Traditional manual methods cannot work satisfactorily for the identification of great number of wake vortex data with high accuracy. Fortunately, the LiDAR intensity data can be explained by integrating LiDAR products with the strategies of computer vision. To overcome the limitation of traditional manual methods, this paper is aimed at developing an automatic method to identify a given set of wake vortices from various aircrafts. The main innovation works are outlined as follows. (1) From the wake vortex data that consists of various aircrafts measured by Wind3D 6000 LiDAR, a grayscale dataset of wake flow is constructed to boost the deep learning model for identifying aircraft wake vortex. (2) Following this, we propose a new method for the identification of aircraft wake vortex by modifying the VGG16 network, providing binary classifications of uncertain behavior patterns for wake vortices. To evaluate the proposed identification model, performance evaluation was conducted on our dataset, where experimental results revealed the values of 0.984, 0.951, 0.959, and 0.955 in terms of accuracy, precision, recall, and F1-score, respectively.