Yi Yang

Mr. Yi Yang


Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 April 2022 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
YYYi Yang
Yi Yang
DZDaoye Zhu
Daoye Zhu
TQTengteng Qu
Tengteng Qu
QWQiangyu Wang
Qiangyu Wang
FRFuhu Ren
Fuhu Ren
CCChengqi Cheng
Chengqi Cheng
In this article, we propose an efficient and generalizable framework based on a deep convolutional neural network (CNN) for multisource remote sensing (RS) data joint classification. While recent methods are mostly based on multistream architectures, we use group convolution (GConv) to construct equivalent network architectures efficiently within a single-stream network. Based on a recent technique called dynamic grouping convolution (DGConv), we further propose a network module named separable DGConv (SepDGConv), to make GConv hyperparameters, and, thus, the overall network architecture, learnable during network training. In the experiments, the proposed method is applied to residual network (ResNet) and UNet, and the adjusted networks are verified on three very diverse benchmark datasets (i.e., Houston2018 data, Berlin data, and MUUFL Gulfport Hyperspectral and LiDAR Airborne Data Set (MUUFL) data). Experimental results demonstrate the effectiveness of the proposed single-stream CNNs, and in particular, SepG-ResNet18 improves the state-of-the-art classification overall accuracy (OA) on hyperspectral–synthetic aperture radar (HS–SAR) Berlin dataset from 62.23% to 68.21%. In the experiments we have two interesting findings. First, using DGConv generally reduces test OA variance. Second, multistream is harmful to model performance if imposed to the first few layers, but becomes beneficial if applied to deeper layers. Altogether, the findings imply that the multistream architecture, instead of being a strictly necessary component in deep learning models for multisource RS data, essentially plays the role of model regularizer. Our code is publicly available at https://github.com/yyyyangyi/CNNs-for-Multi-Source-Remote-Sensing-Data-Fusion . We hope our work can inspire novel research in the future.

29 August 2021 in Remote Sensing
Verified
Journal Article
Open access
PDF file
DZDaoye Zhu
Daoye Zhu
YYYi Yang
Yi Yang
FRFuhu Ren
Fuhu Ren
SMShunji Murai
Shunji Murai
CCChengqi Cheng
Chengqi Cheng
Min HuangMin Huang
Min Huang
The integration analysis of multi-type geospatial information poses challenges to existing spatiotemporal data organization models and analysis models based on deep learning. For earthquake early warning, this study proposes a novel intelligent spatiotemporal grid model based on GeoSOT (SGMG-EEW) for feature fusion of multi-type geospatial data. This model includes a seismic grid sample model (SGSM) and a spatiotemporal grid model based on a three-dimensional group convolution neural network (3DGCNN-SGM). The SGSM solves the problem concerning that the layers of different data types cannot form an ensemble with a consistent data structure and transforms the grid representation of data into grid samples for deep learning. The 3DGCNN-SGM is the first application of group convolution in the deep learning of multi-source geographic information data. It avoids direct superposition calculation of data between different layers, which may negatively affect the deep learning analysis model results. In this study, taking the atmospheric temperature anomaly and historical earthquake precursory data from Japan as an example, an earthquake early warning verification experiment was conducted based on the proposed SGMG-EEW. Five groups of control experiments were designed, namely with the use of atmospheric temperature anomaly data only, use of historical earthquake data only, a non-group convolution control group, a support vector machine control group, and a seismic statistical analysis control group. The results showed that the proposed SGSM is not only compatible with the expression of a single type of spatiotemporal data but can also support multiple types of spatiotemporal data, forming a deep-learning-oriented data structure. Compared with the traditional deep learning model, the proposed 3DGCNN-SGM is more suitable for the integration analysis of multiple types of spatiotemporal data.

7 February 2022 in arXiv
Verified
Preprint
Open access
PDF file
YYYi Yang
Yi Yang
DZDaoye Zhu
Daoye Zhu
TQTengteng Qu
Tengteng Qu
QWQiangyu Wang
Qiangyu Wang
FRFuhu Ren
Fuhu Ren
CCChengqi Cheng
Chengqi Cheng
In this paper, we propose an efficient and generalizable framework based on deep convolutional neural network (CNN) for multi-source remote sensing data joint classification. While recent methods are mostly based on multi-stream architectures, we use group convolution to construct equivalent network architectures efficiently within a single-stream network. We further adopt and improve dynamic grouping convolution (DGConv) to make group convolution hyperparameters, and thus the overall network architecture, learnable during network training. The proposed method therefore can theoretically adjust any modern CNN models to any multi-source remote sensing data set, and can potentially avoid sub-optimal solutions caused by manually decided architecture hyperparameters. In the experiments, the proposed method is applied to ResNet and UNet, and the adjusted networks are verified on three very diverse benchmark data sets (i.e., Houston2018 data, Berlin data, and MUUFL data). Experimental results demonstrate the effectiveness of the proposed single-stream CNNs, and in particular ResNet18-DGConv improves the state-of-the-art classification overall accuracy (OA) on HS-SAR Berlin data set from $62.23\%$ to $68.21\%$. In the experiments we have two interesting findings. First, using DGConv generally reduces test OA variance. Second, multi-stream is harmful to model performance if imposed to the first few layers, but becomes beneficial if applied to deeper layers. Altogether, the findings imply that multi-stream architecture, instead of being a strictly necessary component in deep learning models for multi-source remote sensing data, essentially plays the role of model regularizer. Our code is publicly available at https://github.com/yyyyangyi/Multi-source-RS-DGConv. We hope our work can inspire novel research in the future.