Yidan Zhang

As a key task in remote sensing analysis, semantic segmentation of remote sensing images (RSI) underpins many practical applications. Despite its importance, obtaining dense pixel-wise annotations remains labor-intensive and time-consuming. Unsupervised domain adaptation (UDA) offers a promising solution by utilizing knowledge from labeled source domains for unlabeled target domains, yet its effectiveness is often compromised by significant distribution shifts arising from variations in imaging conditions. To address this challenge, we propose a depth-aware adaptation network (DAAN), a novel two-branch network that explicitly leverages complementary depth information from a digital surface model (DSM) to enhance cross-domain remote sensing segmentation. Unlike conventional UDA methods that primarily focus on semantic features, DAAN incorporates depth data to build a more generalized feature space. This network introduces three key components: an adaptive feature aggregator (AFA) for progressive semantic-depth feature fusion, a gated prediction selection unit (GPSU) that selectively integrates predictions to mitigate the impact of noisy depth measurements, and misalignment-focused residual refinement (MFRR) module that emphasizes poorly aligned target regions during training. Experiments on the ISPRS and GAMUS datasets demonstrate the effectiveness of the proposed method. In particular, DAAN achieves an mIoU of 50.53% and an F1 score of 65.75% for cross-domain segmentation on ISPRS to GAMUS, outperforming models without depth information by 9.17% and 8.99%, respectively. These results demonstrate the advantage of integrating auxiliary geometric information to improve model generalization on unlabeled remote sensing datasets, contributing to higher mapping accuracy, more reliable automated analysis, and enhanced decision-making support.

Knowledge distillation is one of the most effective methods for improving the performance of lightweight detectors, which is crucial for the development of remote sensing (RS) edge intelligence. However, lightweight detectors exhibit significant missed detection when detecting RS targets with arbitrary orientations and extreme scales. This is due to insufficient discriminative features learned by the lightweight student network under standard training, as well as the lack of a targeted, coordinated sample-optimization design. To address these problems, this paper proposes a knowledge Distillation method based on Missed-sample Reinforcement, named DMR. Specifically, a Cross-scale Missed-sample Enhancement module is designed to transfer the teacher's classification knowledge, thereby bridging the core differences in classification. Based on differences between teacher and student predictions, it dynamically selects missed-detection samples, guiding the student model to focus on them. A Dynamic Local Feature Alignment module is also proposed to distill the local key-feature distribution from the teacher model. It efficiently acquires contextual information from local features, guiding the student model to extract discriminative target features. Furthermore, we introduce a generative adversarial process to constrain the student's angle predictions, enabling it to fully utilize the angular information predicted by the teacher model. Extensive experiments on three public RS datasets achieve state-of-the-art (SOTA) performance, demonstrating the effectiveness of the proposed method.

Domain incremental learning for semantic segmentation has gained lots of attention due to its importance for many fields including urban planning and autonomous driving. The catastrophic forgetting problem caused by domain shift has been alleviated by structure expansion of the model or data rehearsal. However, these methods ignore similar contextual knowledge between the new and the old data domain and assume that new knowledge and old knowledge are completely mutually exclusive, which cause the model to be trained in a suboptimal direction. Motivated by the prompt learning, we proposed a new domain incremental learning framework named RS-VDP. The key innovation of RS-VDP is to utilize a visual domain prompt to change the optimization direction from input data space and feature space. First, we designed a domain prompt based on a dynamic location module, which applied a visual domain prompt according to a local entropy map to update the distribution of the input images. Second, in order to filter the feature vectors with high confidence, a representation feature alignment based on an entropy map module is proposed. This module ensures the accuracy and stability of the feature vectors involved in the regularization loss, alleviating the problem of semantic drift. Finally, we introduced a new evaluation metric to measure the overall performance of the incremental learning models, solving the problem that the traditional evaluation metric is affected by the single-task accuracy. Comprehensive experiments demonstrated the effectiveness of the proposed method by significantly reducing the degree of catastrophic forgetting.

无人机凭借全天候、全天时探测优势，在边境监控及灾害救援等领域发挥重要作用。然而，现有的红外-可见光（IR-VIS）多模态目标检测研究过度依赖理想化的“像素级对齐”假设，且数据集普遍存在目标尺度分布极化、类别同质化等问题，导致现有算法在处理具有真实视差及动态尺度演变的无人机航拍数据时，极易出现特征失配、定位漂移及漏检误识。基于以上问题，文中构建了一个面向真实弱配准场景的多模态、多尺度、多类别无人机目标检测基准数据集DIV（drone-based IR-VIS object detection）。该数据集保留了红外与可见光图像之间因无人机传感器安装差异、视角变化及飞行抖动等因素在实际应用中产生的非线性空间偏移。数据集内容涵盖从像素占比极低的微小型目标到显著区域的大型目标，并引入了行人、非机动车、各类车辆等多样化类别。拍摄环境覆盖城市、山区及乡村，并细分为日间、傍晚和弱光三种典型光照场景。同时，通过独立模态的人工高精度标注及多轮交叉验证机制，确保了弱对齐约束下的语义一致性。选取了9种主流多模态目标检测算法在所提数据集上进行基准测试。实验结果表明，在理想对齐数据集中表现优异的方法，在本数据集的弱配准场景下性能出现不同程度的下滑。一些模型方法在特定维度上表现出色，但在应对极端复杂环境时的综合感知能力仍显不足。面对真实环境中的复合挑战，构建一套多维协同优化体系以强化对多模态信息的提取与整合能力，是提升无人机平台感知鲁棒性的关键。提出的数据集有效弥合了多模态探测算法的“理想假设”与无人机数据“现实分布”之间的鸿沟，为构建多维协同优化体系提供了真实的验证平台，在数据层面为弱对齐约束下的跨模态特征融合与鲁棒检测研究提供了关键支撑。DIV数据集发布地址为：https：//www.scidb.cn/preview？dataSetId=d1e8909592e04cc2a1095e62f579ead1&version=V1。