Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval
The comprehensiveness of the raw input data and the effectiveness of feature engineering are two key factors affecting the performance of machine learning. To improve the data comprehensiveness for Global Navigation Satellite System Reflectometry (GNSS-R) ocean wind speed retrieval, this article int...
Published in: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
---|---|
Main Authors: | , , , , , , , |
Format: | Report |
Language: | English |
Published: |
2020
|
Subjects: | |
Online Access: | http://ir.nssc.ac.cn/handle/122/7580 https://doi.org/10.1109/JSTARS.2020.3010879 |
id |
ftchinacadscnssc:oai:ir.nssc.ac.cn:122/7580 |
---|---|
record_format |
openpolar |
spelling |
ftchinacadscnssc:oai:ir.nssc.ac.cn:122/7580 2023-05-15T18:18:35+02:00 Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval WOS:000577878900004 Chu, Xiaohan He, Jie Song, Hongqing Qi, Yue Sun, Yueqiang Bai, Weihua Li, Wei Wu, Qiwu 2020 http://ir.nssc.ac.cn/handle/122/7580 https://doi.org/10.1109/JSTARS.2020.3010879 英语 eng IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING http://ir.nssc.ac.cn/handle/122/7580 doi:10.1109/JSTARS.2020.3010879 null Feature extraction Wind speed Data models Sea surface Sea measurements Machine learning Delay-Doppler map (DDM) GNSS-reflectometry (GNSS-R) multimodal deep learning ocean surface wind retrieval SEA-ICE SCATTERING 期刊论文 2020 ftchinacadscnssc https://doi.org/10.1109/JSTARS.2020.3010879 2021-01-08T01:07:04Z The comprehensiveness of the raw input data and the effectiveness of feature engineering are two key factors affecting the performance of machine learning. To improve the data comprehensiveness for Global Navigation Satellite System Reflectometry (GNSS-R) ocean wind speed retrieval, this article introduces a new input data structure, which is composed of Delay-Doppler maps (DDM) and all satellite receiver status (SRS) parameters. Then, to overcome the difficulty of handcrafted feature engineering and effectively fusion the information of DDM and SRS, we presented a heterogeneous multimodal deep learning (HMDL) method to retrieve the wind speed according to the heterogeneity of the input data. The proposed model is verified by the performance evaluation of realistic data sets obtained from TDS-1. The new input data structure improves the prediction accuracy at 13.5% to 30.7% on mean absolute error (MAE) at 10.6% to 29.5% on the root mean square error (RMSE). The HMDL improves the prediction accuracy at 7.7% on MAE and 7.1% on RMSE. The whole proposed solution improves the prediction accuracy at 36.3% on MAE and 36.8% on RMSE, comparing with the traditional neural network-based solution. The results clearly show that both the introduction of the new input data structure and HMDL effectively improve the accuracy and robustness of GNSS-R wind speed retrieval. Report Sea ice National Space Science Center: NSSC OpenIR (Chinese Academy of Sciences) IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 13 5971 5981 |
institution |
Open Polar |
collection |
National Space Science Center: NSSC OpenIR (Chinese Academy of Sciences) |
op_collection_id |
ftchinacadscnssc |
language |
English |
topic |
Feature extraction Wind speed Data models Sea surface Sea measurements Machine learning Delay-Doppler map (DDM) GNSS-reflectometry (GNSS-R) multimodal deep learning ocean surface wind retrieval SEA-ICE SCATTERING |
spellingShingle |
Feature extraction Wind speed Data models Sea surface Sea measurements Machine learning Delay-Doppler map (DDM) GNSS-reflectometry (GNSS-R) multimodal deep learning ocean surface wind retrieval SEA-ICE SCATTERING Chu, Xiaohan He, Jie Song, Hongqing Qi, Yue Sun, Yueqiang Bai, Weihua Li, Wei Wu, Qiwu Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
topic_facet |
Feature extraction Wind speed Data models Sea surface Sea measurements Machine learning Delay-Doppler map (DDM) GNSS-reflectometry (GNSS-R) multimodal deep learning ocean surface wind retrieval SEA-ICE SCATTERING |
description |
The comprehensiveness of the raw input data and the effectiveness of feature engineering are two key factors affecting the performance of machine learning. To improve the data comprehensiveness for Global Navigation Satellite System Reflectometry (GNSS-R) ocean wind speed retrieval, this article introduces a new input data structure, which is composed of Delay-Doppler maps (DDM) and all satellite receiver status (SRS) parameters. Then, to overcome the difficulty of handcrafted feature engineering and effectively fusion the information of DDM and SRS, we presented a heterogeneous multimodal deep learning (HMDL) method to retrieve the wind speed according to the heterogeneity of the input data. The proposed model is verified by the performance evaluation of realistic data sets obtained from TDS-1. The new input data structure improves the prediction accuracy at 13.5% to 30.7% on mean absolute error (MAE) at 10.6% to 29.5% on the root mean square error (RMSE). The HMDL improves the prediction accuracy at 7.7% on MAE and 7.1% on RMSE. The whole proposed solution improves the prediction accuracy at 36.3% on MAE and 36.8% on RMSE, comparing with the traditional neural network-based solution. The results clearly show that both the introduction of the new input data structure and HMDL effectively improve the accuracy and robustness of GNSS-R wind speed retrieval. |
format |
Report |
author |
Chu, Xiaohan He, Jie Song, Hongqing Qi, Yue Sun, Yueqiang Bai, Weihua Li, Wei Wu, Qiwu |
author_facet |
Chu, Xiaohan He, Jie Song, Hongqing Qi, Yue Sun, Yueqiang Bai, Weihua Li, Wei Wu, Qiwu |
author_sort |
Chu, Xiaohan |
title |
Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
title_short |
Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
title_full |
Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
title_fullStr |
Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
title_full_unstemmed |
Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval |
title_sort |
multimodal deep learning for heterogeneous gnss-r data fusion and ocean wind speed retrieval |
publishDate |
2020 |
url |
http://ir.nssc.ac.cn/handle/122/7580 https://doi.org/10.1109/JSTARS.2020.3010879 |
genre |
Sea ice |
genre_facet |
Sea ice |
op_relation |
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING http://ir.nssc.ac.cn/handle/122/7580 doi:10.1109/JSTARS.2020.3010879 |
op_rights |
null |
op_doi |
https://doi.org/10.1109/JSTARS.2020.3010879 |
container_title |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
container_volume |
13 |
container_start_page |
5971 |
op_container_end_page |
5981 |
_version_ |
1766195218277728256 |