Jian Song

Dr. Jian Song

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

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

Representative Publications

View All Publications
20 June 2024 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
HCHongruixuan Chen
Hongruixuan Chen
JSJian Song
Jian Song
CHChengxi Han
Chengxi Han
Junshi XiaJunshi Xia
Junshi Xia
Naoto YokoyaNaoto Yokoya
Naoto Yokoya
Convolutional neural networks (CNN) and Transformers have made impressive progress in the field of remote sensing change detection (CD). However, both architectures have inherent shortcomings: CNN are constrained by a limited receptive field that may hinder their ability to capture broader spatial contexts, while Transformers are computationally intensive, making them costly to train and deploy on large datasets. Recently, the Mamba architecture, based on state space models, has shown remarkable performance in a series of natural language processing tasks, which can effectively compensate for the shortcomings of the above two architectures. In this paper, we explore for the first time the potential of the Mamba architecture for remote sensing CD tasks. We tailor the corresponding frameworks, called MambaBCD, MambaSCD, and MambaBDA, for binary change detection (BCD), semantic change detection (SCD), and building damage assessment (BDA), respectively. All three frameworks adopt the cutting-edge Visual Mamba architecture as the encoder, which allows full learning of global spatial contextual information from the input images. For the change decoder, which is available in all three architectures, we propose three spatio-temporal relationship modeling mechanisms, which can be naturally combined with the Mamba architecture and fully utilize its attribute to achieve spatio-temporal interaction of multi-temporal features, thereby obtaining accurate change information. On five benchmark datasets, our proposed frameworks outperform current CNN- and Transformer-based approaches without using any complex training strategies or tricks, fully demonstrating the potential of the Mamba architecture in CD tasks. Further experiments show that our architecture is quite robust to degraded data. The source code is available in https://github.com/ChenHongruixuan/MambaCD.

7 November 2023 in ISPRS Journal of Photogrammetry and Remote Sensing
Verified
Journal Article
Closed access
HCHongruixuan Chen
Hongruixuan Chen
JSJian Song
Jian Song
CWChen Wu
Chen Wu
BDBo Du
Bo Du
Naoto YokoyaNaoto Yokoya
Naoto Yokoya

3 June 2025 in Precision Agriculture
Verified
Journal Article
Open access
PDF file
FZFan Zhao
Fan Zhao
YHYinyin He
Yinyin He
JSJian Song
Jian Song
JWJiaqi Wang
Jiaqi Wang
DXDianhan Xi
Dianhan Xi
KMKatsunori Mizuno
Katsunori Mizuno
Purpose: Precise segmentation of blueberry maturity is critical for optimizing harvestschedules and maintaining product quality. Traditional methods, which rely on manualinspection, are not only labor-intensive but also cost-inefficient. This study presents a novelframework that integrates deep learning-based super-resolution reconstruction (SRR) withsemantic segmentation to provide a fast and accurate solution for maturity assessment. Methods: The SRR module enhances image resolution, enabling more detailed feature extraction.Semantic segmentation models—incorporating convolutional neural networks (CNNs),Transformer-based models, and the Mamba-based state space architecture—further improvesegmentation precision. Results: Experimental results indicate that the MambaIR modelachieves a structural similarity index measure (SSIM) of 82.26% in SRR tasks, while the Mamba-based segmentation model attains a mean Intersection over Union (mIoU) of 83.15%. Conclusion: By uniting SRR and semantic segmentation, our framework not only advances thetechnical accuracy of maturity detection but also holds strong potential for real-time, cost-effective deployment in precision agriculture systems, supporting intelligent decision-making at scale.

6 June 2024 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
HCHongruixuan Chen
Hongruixuan Chen
CLCuiling Lan
Cuiling Lan
JSJian Song
Jian Song
CBClifford Broni-Bediako
Clifford Broni-Bediako
Junshi XiaJunshi Xia
Junshi Xia
Naoto YokoyaNaoto Yokoya
Naoto Yokoya
Optical high-resolution imagery and OpenStreetMap (OSM) data are two important data sources of land-cover change detection (CD). Previous related studies focus on utilizing the information in OSM data to aid the CD on optical high-resolution images. This paper pioneers the direct detection of land-cover changes utilizing paired OSM data and optical imagery, thereby expanding the scope of CD tasks. To this end, we propose an object-guided Transformer (ObjFormer) by naturally combining the object-based image analysis (OBIA) technique with the advanced vision Transformer architecture. This combination can significantly reduce the computational overhead in the self-attention module without adding extra parameters or layers. Specifically, ObjFormer has a hierarchical pseudo-siamese encoder consisting of object-guided self-attention modules that extracts multi-level heterogeneous features from OSM data and optical images; a decoder consisting of object-guided cross-attention modules can recover land-cover changes from the extracted heterogeneous features. Beyond basic binary change detection (BCD), this paper raises a new semi-supervised semantic change detection (SCD) task that does not require any manually annotated land-cover labels to train semantic change detectors. Two lightweight semantic decoders are added to ObjFormer to accomplish this task efficiently. A converse cross-entropy (CCE) loss is designed to fully utilize negative samples, contributing to the great performance improvement in this task. A large-scale benchmark dataset called OpenMapCD containing 1,287 map-image pairs covering 40 regions on six continents is constructed to conduct detailed experiments. The results show the effectiveness of our methods in this new kind of CD task. Additionally, case studies in two Japanese cities demonstrate the framework’s generalizability and practical potential.

17 December 2025 in IEEE Geoscience and Remote Sensing Magazine
Verified
Journal Article
Closed access
Junshi XiaJunshi Xia
Junshi Xia
HCHongruixuan Chen
Hongruixuan Chen
CBClifford Broni-Bediako
Clifford Broni-Bediako
YWYimin Wei
Yimin Wei
JSJian Song
Jian Song
Naoto YokoyaNaoto Yokoya
Naoto Yokoya
High-resolution land cover mapping is vital for global challenges such as urban planning, environmental monitoring, disaster response, and sustainable development. Yet, building accurate large-scale datasets remains difficult due to complex terrain, multimodal sensor data, and varying atmospheric conditions. Synthetic aperture radar (SAR), with its cloud-penetrating, all-weather, and day-and-night imaging capabilities, provides a unique advantage, but progress has been hindered by the lack of tailored benchmark datasets. In this article, we introduce OpenEarthMap-SAR, a large-scale benchmark dataset for high-resolution land cover mapping using SAR imagery. OpenEarthMap-SAR includes 1.5 million segments across 5,033 aerial and satellite images (1,024 × 1,024 pixels) from 35 regions in Japan, France, and the United States. It features partially manually annotated and fully pseudo-labeled eight-class land cover labels at a ground sampling distance (GSD) of 0.15–0.5 m. We evaluate state-of-the-art semantic segmentation, unsupervised domain adaptation (UDA), and Image2Image/Label2Image translation methods and propose challenging benchmark settings to foster future development. OpenEarthMap-SAR also serves as the official dataset for track 1 of the IEEE Geoscience and Remote Sensing Society (GRSS) Data Fusion Contest, specifically for the task of semantic segmentation. The dataset and source codes with pretrained models have been made publicly available at https://zenodo.org/records/14622048 and https://github.com/cliffbb/OpenEarthMap-SAR.

1 November 2024 in IEEE Geoscience and Remote Sensing Letters
Verified
Journal Article
Closed access
CBClifford Broni-Bediako
Clifford Broni-Bediako
Junshi XiaJunshi Xia
Junshi Xia
JSJian Song
Jian Song
HCHongruixuan Chen
Hongruixuan Chen
MSMennatullah Siam
Mennatullah Siam
Naoto YokoyaNaoto Yokoya
Naoto Yokoya
Learning with limited labelled data is a challenging problem in various applications, including remote sensing. Few-shot semantic segmentation is one approach that can encourage deep learning models to learn from few labelled examples for novel classes not seen during the training. The generalized few-shot segmentation setting has an additional challenge which encourages models not only to adapt to the novel classes but also to maintain strong performance on the training base classes. While previous datasets and benchmarks discussed the few-shot segmentation setting in remote sensing, we are the first to propose a generalized few-shot segmentation benchmark for remote sensing. The generalized setting is more realistic and challenging, which necessitates exploring it within the remote sensing context. We release the dataset augmenting OpenEarthMap with additional classes labelled for the generalized few-shot evaluation setting. The dataset is released during the OpenEarthMap land cover mapping generalized few-shot challenge in the L3D-IVU workshop in conjunction with CVPR 2024. In this work, we summarize the dataset and challenge details in addition to providing the benchmark results on the two phases of the challenge for the validation and test sets.