Most of the current malware detection methods running on Android are based on signature and cloud technologies leading to poor protection against new types of malware. Deep learning techniques take Android malware detection to a new level. Still, most deep learning-based Android malware detection methods are too inefficient or even unworkable on Android devices due to their high resource consumption. Therefore, this paper proposes MSFDroid, a lightweight multi-source fast Android malware detection model, which uses information from the internal files of the Android application package in several dimensions to build base models for ensemble learning. Meanwhile, this paper proposes an adaptive soft voting method by dynamically adjusting the weights of each base model to overcome the noise generated by traditional soft voting and thus improves the performance. It also proposes adaptive shrinkage convolutional unit that can dynamically adjust the convolutional kernel’s weight and the activation function’s threshold to improve the expressiveness of the CNN. The proposed method is tested on public datasets and on several real devices. The experimental results show that it achieves a better trade-off between performance and efficiency by significantly improving the detection speed while achieving a comparable performance compared to other deep learning methods.

Surface defect detection is vital in industrial quality control. While deep learning has largely automated inspection, accurately locating defects with large-scale variations or those difficult to distinguish from similar backgrounds remains challenging. Furthermore, achieving high-precision and real-time performance under limited computational resources in deployment environments complicates effective solutions. In this work, we propose SD-GASNet, a network based on a self-distillation model compression strategy. To identify subtle defects, we design an Alignment, Enhancement, and Synchronization Feature Pyramid Network (AES-FPN) fusion network incorporating the Frequency Domain Information Gathering-and-Allocation (FIGA) mechanism and the Channel Synchronization (CS) module for industrial images from different sensors. Specifically, FIGA refines features via the Multi-scale Feature Alignment (MFA) module, then the Frequency-Guided Perception Enhancement Module (FGPEM) extracts high- and low-frequency information to enhance spatial representation. The CS module compensates for information loss during feature fusion. Addressing computational constraints, we adopt self-distillation with an Enhanced KL divergence loss function to boost lightweight model performance. Extensive experiments on three public datasets (NEU-DET, PCB, and TILDA) demonstrate that SD-GASNet achieves state-of-the-art performance with excellent generalization, delivering superior accuracy and a competitive inference speed of 180 FPS, offering a robust and generalizable solution for sensor-based industrial imaging applications.