With the rapid development of artificial intelligence technology in the field of intelligent manufacturing, convolutional neural networks (CNNs) have shown excellent performance and generalization capabilities in industrial applications. However, the huge computational and resource requirements of CNNs have brought great obstacles to their deployment on low-end hardware platforms. To address this issue, this paper proposes a scalable CNN accelerator that can operate on low-performance Field-Programmable Gate Arrays (FPGAs), which is aimed at tackling the challenge of efficiently running complex neural network models on resource-constrained hardware platforms. This study specifically optimizes depthwise separable convolution and the squeeze-and-excite module to improve their computational efficiency. The proposed accelerator allows for the flexible adjustment of hardware resource consumption and computational speed through configurable parameters, making it adaptable to FPGAs with varying performance and different application requirements. By fully exploiting the characteristics of depthwise separable convolution, the accelerator optimizes the convolution computation process, enabling flexible and independent module stackings at different stages of computation. This results in an optimized balance between hardware resource consumption and computation time. Compared to ARM CPUs, the proposed approach yields at least a 1.47× performance improvement, and compared to other FPGA solutions, it saves over 90% of Digital Signal Processors (DSPs). Additionally, the optimized computational flow significantly reduces the accelerator’s reliance on internal caches, minimizing data latency and further improving overall processing efficiency.

Few-shot image classification seeks to recognize novel categories from only a handful of labeled examples, but conventional metric-based methods that rely mainly on global image features often produce unstable prototypes under extreme data scarcity, while local-descriptor approaches can lose context and suffer from inter-class local-pattern overlap. To address these limitations, we propose a Global–Local Feature Fusion network that combines a frozen, pretrained global feature branch with a self-attention based multi-local feature fusion branch. Multiple random crops are encoded by a shared backbone (ResNet-12), projected to Query/Key/Value embeddings, and fused via scaled dot-product self-attention to suppress background noise and highlight discriminative local cues. The fused local representation is concatenated with the global feature to form robust class prototypes used in a prototypical-network style classifier. On four benchmarks, our method achieves strong improvements: Mini-ImageNet 70.31% ± 0.20 (1-shot)/85.91% ± 0.13 (5-shot), Tiered-ImageNet 73.37% ± 0.22/87.62% ± 0.14, FC-100 47.01% ± 0.20/64.13% ± 0.19, and CUB-200-2011 82.80% ± 0.18/93.19% ± 0.09, demonstrating consistent gains over competitive baselines. Ablation studies show that (1) naive local averaging improves over global-only baselines, (2) self-attention fusion yields a large additional gain (e.g., +4.50% in 1-shot on Mini-ImageNet), and (3) concatenating global and fused local features gives the best overall performance. These results indicate that explicitly modeling inter-patch relations and fusing multi-granularity cues produces markedly more discriminative prototypes in few-shot regimes.

The rapid development of network communication technology has led to an increased focus on the security of image storage and transmission in multimedia information. This paper proposes an enhanced image security communication scheme based on Fibonacci interleaved diffusion and non-degenerate chaotic system to address the inadequacy of current image encryption technology. The scheme utilizes a hash function to extract the hash characteristic values of the plaintext image, generating initial perturbation keys to drive the chaotic system to generate initial pseudo-random sequences. Subsequently, the input image is subjected to a light scrambling process at the bit level. The Q matrix generated by the Fibonacci sequence is then employed to diffuse the obtained intermediate cipher image. The final ciphertext image is then generated by random direction confusion. Throughout the encryption process, plaintext correlation mechanisms are employed. Consequently, due to the feedback loop of the plaintext, this algorithm is capable of resisting known-plaintext attacks and chosen-plaintext attacks. Theoretical analysis and empirical results demonstrate that the algorithm fulfils the cryptographic requirements of confusion, diffusion, and avalanche effects, while also exhibiting a robust password space and excellent numerical statistical properties. Consequently, the security enhancement mechanism based on Fibonacci interleaved diffusion and non-degenerate chaotic system proposed in this paper effectively enhances the algorithm’s resistance to cryptographic attacks.

With the rapid development of big data and artificial intelligence, the problem of image privacy leakage has become increasingly prominent, especially for images containing sensitive information such as faces, which poses a higher security risk. In order to improve the security and efficiency of image privacy protection, this paper proposes an image encryption scheme that integrates face detection and multi-level encryption technology. Specifically, a multi-task convolutional neural network (MTCNN) is used to accurately extract the face area to ensure accurate positioning and high processing efficiency. For the extracted face area, a hierarchical encryption framework is constructed using chaotic systems, lightweight block permutations, RNA cryptographic systems, and bit diffusion, which increases data complexity and unpredictability. In addition, a key update mechanism based on dynamic feedback is introduced to enable the key to change in real time during the encryption process, effectively resisting known plaintext and chosen plaintext attacks. Experimental results show that the scheme performs well in terms of encryption security, robustness, computational efficiency, and image reconstruction quality. This study provides a practical and effective solution for the secure storage and transmission of sensitive face images, and provides valuable support for image privacy protection in intelligent systems.