Mofan Dai

Mofan Dai

Auto-generated profile, based on publicly available information

Mofan Dai

School of Surveying and Mapping, PLA Information Engineering University, Zhengzhou, China

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
9 February 2024 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
MDMofan Dai
Mofan Dai
SXShuai Xing
Shuai Xing
QXQing Xu
Qing Xu
PLPengcheng Li
Pengcheng Li
JPJiechen Pan
Jiechen Pan
HWHanyun Wang
Hanyun Wang
Existing airborne laser scanning (ALS) point cloud semantic segmentation approaches are limited by their overreliances on sufficient point-wise annotations that further confine their generalization ability to new scenes. To overcome these problems, a novel three-stage multi-prototype relational network (Thr-MPRNet) is proposed for few-shot ALS point cloud semantic segmentation by transferring knowledge from well-annotated photogrammetric point clouds. In MPRNet, a 3D few-shot learning structure containing a feature learner and a relation learner is built to learn meta-knowledge from multiple point-wise tasks, and a multi-prototype generator is designed to represent the semantic distribution of point clouds that can dynamically adapt to large-scale scenarios. Then, to transfer knowledge across different domains, MPRNet is trained in a unified framework with three task-based learning stages. Prior knowledge is first meta-learned from the source photogrammetric point clouds and then transferred to novel target datasets with a few labeled ALS point clouds. Finally, the MPRNet can be flexibly generalized to the unlabeled target ALS point clouds without further retraining from scratch. In the experiments, the SensatUrban dataset is used as the source photogrammetric point clouds, and two ALS point cloud datasets (ISPRS and DALES) are used to evaluate the few-shot semantic segmentation ability of the proposed method. The experiments demonstrate that Thr-MPRNet obtains promising generalization performance on different target datasets. More importantly, it outperforms supervised networks with 10% labeled samples. In summary, the proposed method achieves state-of-the-art cross-domain semantic segmentation performance and greatly alleviates the dependence on ALS point cloud annotations.

30 December 2023 in International Journal of Applied Earth Observation and Geoinformation
Verified
Journal Article
Open access
MDMofan Dai
Mofan Dai
SXShuai Xing
Shuai Xing
QXQing Xu
Qing Xu
HWHanyun Wang
Hanyun Wang
PLPengcheng Li
Pengcheng Li
YLYuqiong Li
Yuqiong Li

29 August 2025 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Open access
Guoping ZhangGuoping Zhang
Guoping Zhang
MGMing Gao
Ming Gao
SXShuai Xing
Shuai Xing
RKRuiyao Kong
Ruiyao Kong
MDMofan Dai
Mofan Dai
QXQing Xu
Qing Xu
Although the supraglacial lakes are significant for a comprehensive understanding of the glacier material balance mechanism, obtaining accurate supraglacial lake data by satellite remote sensing is challenging. The lack of three-dimensional information limits the accuracy of image-based methods, while the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) data are limited to satellite trajectories. In this study, the detection and mapping of the supraglacial lakes are implemented by jointly processing the ICESat-2 and Sentinel-2 data. The lake-region images are extracted with the help of ICESat-2 photons, and the lake depths are retrieved. The depths of four supraglacial lakes on the Amery Ice Shelf (AIS), Antarctica, were mapped by trained machine learning methods and Bayesian Optimization (BO). The performance of the proposed method is evaluated by comparing it with manual results or self-verification. The results show that joint lake region extraction accuracy is 95.16%. The proposed photon extraction method, combined with BO-CatBoost, achieves the highest accuracy in retrieval, with an RMSE of 0.17 m. This study enables automatic operation without input parameters, thus improving the performance and automation level of supraglacial lake observation. With more abundant ICESat-2 data and multispectral images, the proposed method can automatically obtain large-scale surface melting in time and space.

30 December 2024 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
MDMofan Dai
Mofan Dai
SXShuai Xing
Shuai Xing
QXQing Xu
Qing Xu
PLPengcheng Li
Pengcheng Li
JPJiechen Pan
Jiechen Pan
HWHanyun Wang
Hanyun Wang
Feature learning of airborne laser scanning (ALS) point clouds is challenged by both the limited annotated samples and imbalanced class distribution. An intuitive way involves pretraining on a well-annotated source dataset and finetuning on a limited target dataset. However, cross-domain challenges such as heterogeneous point cloud density, varying terrain features, and inconsistent object categories complicate transfer learning for 3D land cover classification. In this paper, we address these issues by separating the cross-domain ALS point cloud semantic segmentation into two subsequent subtasks, i.e., the cross-domain transfer learning subtask and the intra-domain class-incremental learning subtask and we use a well-annotated photogrammetric point cloud dataset as the source dataset. To mitigate domain discrepancies, the first subtask employs domain-adversarial training to learn from base categories that are shared between source and target datasets. Then, the second subtask incrementally learns new categories that are specific within the target dataset using an incremental feature-semantic distillation module and a semantic adversarial learning module, while retaining base category knowledge. Experimental results evaluated on three ALS point cloud datasets (ISPRS, DALES, and H3D) with different semantics show state-of-the-art cross-domain performance with few labeled samples. Compared with few-shot learning methods, our method shows promising generalization ability particularly on domain-specific categories, greatly alleviating the dependence on ALS point cloud annotations.

16 April 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
JPJiechen Pan
Jiechen Pan
SXShuai Xing
Shuai Xing
WSWei Sun
Wei Sun
MDMofan Dai
Mofan Dai
DWDandi Wang
Dandi Wang
LZLu Zhi
Lu Zhi
Aerial point cloud semantic segmentation technology enables intelligent interpretation of three-dimensional geographic information with precise structural representations and surface features. Recent Transformer-based point cloud semantic segmentation methods using multi-layer self-attention mechanisms can dynamically learn important relationships between points. With sufficient labeled samples, it can effectively capture complex global structures and long-range dependencies within point clouds. However, point-wise annotations of large-scale aerial point clouds are labor-intensive and time-consuming. Under limited annotation conditions, there are critical limitations of Transformer-based point cloud semantic segmentation methods, including inadequate modeling of local structures and ineffective classification of small objects. Therefore, to address the aforementioned challenges, this study proposes a geometric-aware transformer for a weakly supervised semantic segmentation method for aerial point clouds. Firstly, a geometry-aware encoding layer that explicitly incorporates foundational geometric descriptors (linearity, planarity, etc.) to compensate for the lack of fine-grained local supervision, enhancing feature discrimination under sparse labels. And a bidirectional interactive attention module that explicitly models both the influence of local neighborhoods on their center and the subsequent diffusion of global-context-aware semantics from the center back to its neighbors, thereby maximizing the utility of sparse annotations. Secondly, a pseudo-label adaptive training module leveraging abundant unlabeled data is designed to enhance model performance and generalization capability. Experimental results on the SensatUrban and DALES datasets demonstrate that the proposed method achieves 58.69% mIoU on SensatUrban (with 10% labeled data) and 74.80% mIoU on DALES (with 10% labeled data). These results surpass both the fully supervised Point Transformer baseline and multiple early fully supervised methods on both datasets.