Real-time remote sensing segmentation technology is crucial for unmanned aerial vehicles (UAVs) in battlefield surveillance, land characterization observation, earthquake disaster assessment, etc., and can significantly enhance the application value of UAVs in military and civilian fields. To realize this potential, it is essential to develop real-time semantic segmentation methods that can be applied to resource-limited platforms, such as edge devices. The majority of mainstream real-time semantic segmentation methods rely on convolutional neural networks (CNNs) and transformers. However, CNNs cannot effectively capture long-range dependencies, while transformers have high computational complexity. This paper proposes a novel remote sensing Mamba architecture for real-time segmentation tasks in remote sensing, named RTMamba. Specifically, the backbone utilizes a Visual State-Space (VSS) block to extract deep features and maintains linear computational complexity, thereby capturing long-range contextual information. Additionally, a novel Inverted Triangle Pyramid Pooling (ITP) module is incorporated into the decoder. The ITP module can effectively filter redundant feature information and enhance the perception of objects and their boundaries in remote sensing images. Extensive experiments were conducted on three challenging aerial remote sensing segmentation benchmarks, including Vaihingen, Potsdam, and LoveDA. The results show that RTMamba achieves competitive performance advantages in terms of segmentation accuracy and inference speed compared to state-of-the-art CNN and transformer methods. To further validate the deployment potential of the model on embedded devices with limited resources, such as UAVs, we conducted tests on the Jetson AGX Orin edge device. The experimental results demonstrate that RTMamba achieves impressive real-time segmentation performance.

Small object detection (SOD) in remote sensing images (RSIs) is a challenging task due to scale variation, severe occlusion, and complex backgrounds, often leading to high miss and false detection rates. To address these issues, this paper proposes a novel detection framework named HSF-YOLO, which is designed to jointly enhance feature encoding, attention interaction, and localization precision within the YOLOv8 backbone. Specifically, we introduce three tailored modules: Hybrid Atrous Enhanced Convolution (HAEC), a Spatial–Interactive–Shuffle attention module (C2f_SIS), and a Focal Gradient Refinement Loss (FGR-Loss). The HAEC module captures multi-scale semantic and fine-grained local information through parallel atrous and standard convolutions, thereby enhancing small object representation across scales. The C2f_SIS module fuses spatial and improved channel attention with a channel shuffle strategy to enhance feature interaction and suppress background noise. The FGR-Loss incorporates gradient-aware localization, focal weighting, and separation-aware constraints to improve regression accuracy and training robustness. Extensive experiments were conducted on three public remote sensing datasets. Compared with the baseline YOLOv8, HSF-YOLO improved mAP@0.5 and mAP@0.5:0.95 by 5.7% and 4.0% on the VisDrone2019 dataset, by 2.3% and 2.5% on the DIOR dataset, and by 2.3% and 2.1% on the NWPU VHR-10 dataset, respectively. These results confirm that HSF-YOLO is a unified and effective solution for small object detection in complex RSI scenarios, offering a good balance between accuracy and efficiency.

Knowledge Tracing aims to assess students’ mastery of knowledge concepts in real time, playing a crucial role in providing personalized learning services in intelligent tutoring systems. In recent years, researchers have attempted to introduce Neural Architecture Search (NAS) into knowledge tracing tasks to automatically design more efficient network structures. However, existing NAS-based methods for Knowledge Tracing suffer from excessively large search spaces and slow search efficiency, which significantly constrain their practical applications. To address these limitations, this paper proposes an Adaptive Neural Architecture Search framework based on Transformers for KT, called ANT-KT. Specifically, we design an enhanced encoder that combines convolution operations with state vectors to capture both local and global dependencies in students’ learning sequences. Moreover, an optimized decoder with a linear attention mechanism is introduced to improve the efficiency of modeling long-term student knowledge state evolution. We further propose an evolutionary NAS algorithm that incorporates a model optimization efficiency objective and a dynamic search space reduction strategy, enabling the discovery of high-performing yet computationally efficient architectures. Experimental results on two large-scale real-world datasets, EdNet and RAIEd2020, demonstrate that ANT-KT significantly reduces time costs across all stages of NAS while achieving performance improvements on multiple evaluation metrics, validating the efficiency and practicality of the proposed method.