Haochen Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Haochen Sun has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Hongping Li
College of Marine Technology, Ocean University of China, Qingdao 266100, China
3 shared publications - Tianyu Xia
Faculty of Information Science and Engineering, College of Marine Technology, Ocean University of China, Qingdao 266100, China
2 shared publications Cuishu Li
1 shared publicationFan Yang
1 shared publication- Hao Yu
College of Marine Technology, Faculty of Information Science and Engineering, Ocean University of China, Qingdao 266100, China
1 shared publication Ming Xu
1 shared publication
Latest Publications
- Open Access
- 9 October 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (19) , 3394
3DV-Unet: Eddy-Resolving Reconstruction of Three-Dimensional Upper-Ocean Physical Fields from Satellite Observations
Three-dimensional (3D) ocean physical fields are essential for understanding ocean dynamics, but reconstructing them solely from sea-surface remote sensing remains challenging. We present 3DV-Unet, an end-to-end deep learning framework that reconstructs eddy-resolving three-dimensional essential ocean variables (temperature, salinity, and currents) from multi-source satellite data. The model employs a 3D Vision Transformer bottleneck to capture cross-depth and cross-variable dependencies, ensuring physically consistent reconstruction. Trained on 2011–2019 reanalysis and satellite data, 3DV-Unet achieves RMSEs of ~0.30 °C for temperature, 0.11 psu for salinity, and 0.05 m/s for currents, with all R2 values above 0.93. Error analyses further indicate higher reconstruction errors in dynamically complex regions such as the Kuroshio Extension, while spectral analysis indicates good agreement at 100 km+ but systematic deviation in the 20–100 km band. Independent validation against 6113 Argo profiles confirms its ability to reproduce realistic vertical thermohaline structures. Moreover, the reconstructed 3D fields capture mesoscale eddy structures and their life cycle, offering a valuable basis for investigating ocean circulation, energy transport, and regional variability. These results demonstrate the potential of end-to-end volumetric deep learning for advancing high-resolution 3D ocean reconstruction and supporting physical oceanography and climate studies.
- 9 October 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (19) , 3394
Open Access3DV-Unet: Eddy-Resolving Reconstruction of Three-Dimensional Upper-Ocean Physical Fields from Satellite Observations
Three-dimensional (3D) ocean physical fields are essential for understanding ocean dynamics, but reconstructing them solely from sea-surface remote sensing remains challenging. We present 3DV-Unet, an end-to-end deep learning framework that reconstructs eddy-resolving three-dimensional essential ocean variables (temperature, salinity, and currents) from multi-source satellite data. The model employs a 3D Vision Transformer bottleneck to capture cross-depth and cross-variable dependencies, ensuring physically consistent reconstruction. Trained on 2011–2019 reanalysis and satellite data, 3DV-Unet achieves RMSEs of ~0.30 °C for temperature, 0.11 psu for salinity, and 0.05 m/s for currents, with all R2 values above 0.93. Error analyses further indicate higher reconstruction errors in dynamically complex regions such as the Kuroshio Extension, while spectral analysis indicates good agreement at 100 km+ but systematic deviation in the 20–100 km band. Independent validation against 6113 Argo profiles confirms its ability to reproduce realistic vertical thermohaline structures. Moreover, the reconstructed 3D fields capture mesoscale eddy structures and their life cycle, offering a valuable basis for investigating ocean circulation, energy transport, and regional variability. These results demonstrate the potential of end-to-end volumetric deep learning for advancing high-resolution 3D ocean reconstruction and supporting physical oceanography and climate studies.
- Open Access
- 23 December 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (24) , 4808
Detecting Ocean Eddies with a Lightweight and Efficient Convolutional Network
As a ubiquitous mesoscale phenomenon, ocean eddies significantly impact ocean energy and mass exchange. Detecting these eddies accurately and efficiently has become a research focus in ocean remote sensing. Many traditional detection methods, rooted in physical principles, often encounter challenges in practical applications due to their complex parameter settings, while effective, deep learning models can be limited by the high computational demands of their extensive parameters. Therefore, this paper proposes a new approach to eddy detection based on the altimeter data, the Ghost Attention Deeplab Network (GAD-Net), which is a lightweight and efficient semantic segmentation model designed to address these issues. The encoder of GAD-Net consists of a lightweight ECA+GhostNet and an Atrous Spatial Pyramid Pooling (ASPP) module. And the decoder integrates an Efficient Attention Network (EAN) module and an Efficient Ghost Feature Integration (EGFI) module. Experimental results show that GAD-Net outperforms other models in evaluation indices, with a lighter model size and lower computational complexity. It also outperforms other segmentation models in actual detection results in different sea areas. Furthermore, GAD-Net achieves detection results comparable to the Py-Eddy-Tracker (PET) method with a smaller eddy radius and a faster detection speed. The model and the constructed eddy dataset are publicly available.
- 23 December 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (24) , 4808
Open AccessDetecting Ocean Eddies with a Lightweight and Efficient Convolutional Network
As a ubiquitous mesoscale phenomenon, ocean eddies significantly impact ocean energy and mass exchange. Detecting these eddies accurately and efficiently has become a research focus in ocean remote sensing. Many traditional detection methods, rooted in physical principles, often encounter challenges in practical applications due to their complex parameter settings, while effective, deep learning models can be limited by the high computational demands of their extensive parameters. Therefore, this paper proposes a new approach to eddy detection based on the altimeter data, the Ghost Attention Deeplab Network (GAD-Net), which is a lightweight and efficient semantic segmentation model designed to address these issues. The encoder of GAD-Net consists of a lightweight ECA+GhostNet and an Atrous Spatial Pyramid Pooling (ASPP) module. And the decoder integrates an Efficient Attention Network (EAN) module and an Efficient Ghost Feature Integration (EGFI) module. Experimental results show that GAD-Net outperforms other models in evaluation indices, with a lighter model size and lower computational complexity. It also outperforms other segmentation models in actual detection results in different sea areas. Furthermore, GAD-Net achieves detection results comparable to the Py-Eddy-Tracker (PET) method with a smaller eddy radius and a faster detection speed. The model and the constructed eddy dataset are publicly available.
- Open Access
- 10 November 2023
- Published by Frontiers Media SA in Frontiers in Marine Science
- Vol. 10, 1266452
A lightweight deep learning model for ocean eddy detection
Ocean eddies are typical oceanic mesoscale phenomena that are numerous, widely distributed and have high energy. Traditional eddy detection methods are mainly based on physical mechanisms with high accuracy. However, the large number of steps and complex parameter settings limit their applicability for most users. With the rapid development of deep learning techniques, object detection models have been broadly used in the field of ocean remote sensing. This paper proposes a lightweight eddy detection model, ghost eddy detection YOLO (GED-YOLO), based on sea level anomaly data and the “You Only Look Once” (YOLO) series models. The proposed model used ECA+GhostNet as the backbone network and an atrous spatial pyramid pooling network as the feature enhancement network. The ghost eddy detection path aggregation network was proposed for feature fusion, which reduced the number of model parameters and improved the detection performance. The experimental results showed that GED-YOLO achieved better detection precision and smaller parameter size. Its mAP was 95.11% and the parameter size was 22.56 MB. In addition, the test experiment results showed that GED-YOLO had similar eddy detection performance and faster detection speed compared to the traditional physical method.
- 10 November 2023
- Published by Frontiers Media SA in Frontiers in Marine Science
- Vol. 10, 1266452
Open AccessA lightweight deep learning model for ocean eddy detection
Ocean eddies are typical oceanic mesoscale phenomena that are numerous, widely distributed and have high energy. Traditional eddy detection methods are mainly based on physical mechanisms with high accuracy. However, the large number of steps and complex parameter settings limit their applicability for most users. With the rapid development of deep learning techniques, object detection models have been broadly used in the field of ocean remote sensing. This paper proposes a lightweight eddy detection model, ghost eddy detection YOLO (GED-YOLO), based on sea level anomaly data and the “You Only Look Once” (YOLO) series models. The proposed model used ECA+GhostNet as the backbone network and an atrous spatial pyramid pooling network as the feature enhancement network. The ghost eddy detection path aggregation network was proposed for feature fusion, which reduced the number of model parameters and improved the detection performance. The experimental results showed that GED-YOLO achieved better detection precision and smaller parameter size. Its mAP was 95.11% and the parameter size was 22.56 MB. In addition, the test experiment results showed that GED-YOLO had similar eddy detection performance and faster detection speed compared to the traditional physical method.