Multimodal perception for student-state monitoring is difficult to deploy in rural classrooms because sensors are noisy and computing resources are highly constrained. This work targets these challenges by enabling noise-resilient, multimodal, real-time student-state recognition on low-cost edge devices. We propose LightEdu-Net, a sensor-noise-adaptive Transformer-based multimodal network that integrates visual, physiological, and environmental signals in a unified lightweight architecture. The model incorporates three key components: a sensor noise adaptive module (SNAM) to suppress degraded sensor inputs, a cross-modal attention fusion module (CMAF) to capture complementary temporal dependencies across modalities, and an edge-aware knowledge distillation module (EAKD) to transfer knowledge from high-capacity teachers to an embedded-friendly student network. We construct a multimodal behavioral dataset from several rural schools and formulate student-state recognition as a multimodal classification task with explicit evaluation of noise robustness and edge deployability. Experiments show that LightEdu-Net achieves 92.4% accuracy with an F1-score of 91.4%, outperforming representative lightweight CNN and Transformer baselines. Under a noise level of 0.3, accuracy drops by only 1.1%, indicating strong robustness to sensor degradation. Deployment experiments further show that the model operates in real time on Jetson Nano with a latency of 42.8 ms (23.4 FPS) and maintains stable high accuracy on Raspberry Pi 4B and Intel NUC platforms. Beyond technical performance, the proposed system provides a low-cost and quantifiable mechanism for capturing fine-grained learning process indicators, offering new data support for educational economics studies on instructional efficiency and resource allocation in underdeveloped regions.

With the rapid development of intelligent sensing systems, digital monitoring platforms, and multisource data acquisition technologies, accurate identification of operational states and potential risks from heterogeneous sensing signals has become an important research issue in artificial intelligence-driven sensing. Existing studies have primarily focused on either textual information understanding or behavioral data analysis, with limited attention paid to jointly modeling the consistency between textual declarations and executed behaviors. As a result, many potential risks that have not yet manifested as significant anomalies but already involve execution deviations are difficult to detect in a timely manner. To address this issue, a language–behavior consistency sensing framework for multisource sensing signals is proposed. Textual sensing signals and behavioral sensing signals are mapped into a shared state logic space, and intelligent perception and quantitative analysis of deviations between textual states and executed states are achieved through a textual state logic extraction module, an observed behavioral state modeling module, and a language–behavior consistency measurement module. Systematic experiments were conducted on a multisource sensing dataset containing public declaration texts, operation reports, behavioral logs, resource allocation records, and state-response information. The results show that the proposed method achieved the best performance in the baseline comparison experiment, with a language–behavior consistency score ($LCS$) of $0.742$, an $AUC$ of $0.846$, an $F1$-score of $0.811$, a Precision of $0.802$, and an explanation consistency score ($ECS$) of $0.821$, clearly outperforming advanced methods such as FinBERT, LSTM, Multimodal Transformer, and the Contrastive Multimodal Model. These results demonstrate that language–behavior consistency sensing can effectively fuse multisource sensing information and improve complex system state identification, anomaly early warning, and risk perception, providing an interpretable artificial intelligence-driven sensing framework with the potential to support industrial operation and maintenance, intelligent manufacturing, digital infrastructure management, and other intelligent monitoring scenarios.

In UAV (unmanned aerial vehicle)-enabled precision agriculture, achieving high-accuracy disease diagnosis and yield estimation simultaneously on resource-constrained edge devices remains a significant challenge. Existing solutions are commonly hindered by conflicts in visual feature scales, the absence of explicit agronomic causal logic, and the trade-off between lightweight design and global modeling capability. To address these challenges, a heterogeneous dual-stream state-space framework termed SymbioMamba is proposed. The proposed framework incorporates three key innovations: first, a heterogeneous dual-stream encoder is constructed, in which a micro-texture stream captures high-frequency disease details while a macro-context-scan stream models field-scale biomass continuity; second, a pathology–biomass collaborative interaction (PBCI) module is designed to explicitly inject the biological prior—disease stress leading to yield reduction—into the feature space. Third, a topology-aligning cross-architecture distillation (TACAD) paradigm is introduced to transfer global knowledge from a heavyweight teacher to a lightweight student. Experimental results from a maize UAV dataset comprising 12,074 annotated image patches demonstrate that SymbioMamba achieves 89.4% mAP@0.5 and an $R^2$ of 0.915. Compared to the industry-standard YOLOv11, the framework improves mAP@0.5:0.95 by 2.4% while reducing the parameter count to 6.2 M—a 50% decrease relative to monolithic state-space baselines. Furthermore, yield prediction error is significantly reduced to an RMSE of 485.6 kg/ha. With a compact model size of 6.2 M parameters and 2.4 G FLOPs, SymbioMamba attains an inference speed of 38.2 FPS on the NVIDIA Jetson AGX Orin platform, providing a high-performance, real-time solution for intelligent agricultural phenotypic analysis.