In this study, series networks (AlexNet and VGG-19) and directed acyclic graph (DAG) networks (ResNet-18, ResNet-50, and ResNet-101) with transfer learning were employed to identify and classify 13 classes of apples from 7439 images. Two training datasets, model evaluation metrics, and three visualization methods were used to objectively assess, compare, and interpret five Convolutional Neural Network (CNN)-based models. The results show that the dataset configuration had a significant impact on the classification results, as all models achieved over 96.1% accuracy on dataset A (training-to-testing = 2.4:1.0) compared to 89.4–93.9% accuracy on dataset B (training-to-testing = 1.0:3.7). VGG-19 achieved the highest accuracy of 100.0% on dataset A and 93.9% on dataset B. Moreover, for networks of the same framework, the model size, accuracy, and training and testing times increased as the model depth (number of layers) increased. Furthermore, feature visualization, strongest activations, and local interpretable model-agnostic explanations techniques were used to show the understanding of apple images by different trained models, as well as to reveal how and why the models make classification decisions. These results improve the interpretability and credibility of CNN-based models, which provides guidance for future applications of deep learning methods in agriculture.

This study investigates the effectiveness of using conditional generative adversarial networks (CGAN) to synthesize Optical Coherence Tomography (OCT) images for medical diagnosis. Specifically, the CGAN model is trained to generate images representing various eye conditions, including normal retina, vitreous warts (DRUSEN), choroidal neovascularization (CNV), and diabetic macular edema (DME), creating a dataset of 102,400 synthetic images per condition. The quality of these images is evaluated using two methods. First, 18 transfer-learning neural networks (including AlexNet, VGGNet16, GoogleNet) assess image quality through model-scoring metrics, resulting in an accuracy rate of 97.4% to 99.9% and an F1 Score of 95.3% to 100% across conditions. Second, interpretative analysis techniques (GRAD-CAM, occlusion sensitivity, LIME) compare the decision score distribution of real and synthetic images, further validating the CGAN network’s performance. The results indicate that CGAN-generated OCT images closely resemble real images and could significantly contribute to medical datasets.

To achieve a non-destructive and rapid detection of oyster freshness, an intelligent method using deep learning fused with malondialdehyde (MDA) and total sulfhydryl groups (SH) information was proposed. In this study, an “MDA-SH-storage days” polynomial fitting model and oyster meat image dataset were first built. AleNet-MDA and AlxNet-SH classification models were then constructed to automatically identify and classify four levels of oyster meat images with overall accuracies of 92.72% and 94.06%, respectively. Next, the outputs of the two models were used as the inputs to “MDA-SH-storage days” model, which ultimately succeeded in predicting the corresponding MDA content, SH content and storage day for an oyster image within 0.03 ms. Furthermore, the interpretability of the two models for oyster meat image were also investigated by feature visualization and strongest activations techniques. Thus, this study brings new thoughts on oyster freshness prediction from the perspective of computer vision and artificial intelligence.

Gabor wavelet has excellent properties in extracting local spatial and frequency domain information of a target. While it cannot inherently constitute an orthogonal basis, it can form a compact frame under specific parameters, which provides the possibility of using Gabor wavelets for feature extraction. In this paper, a novel Gabor wavelet initialized sparse filtering (GWISF) is designed to bolster the interpretability and efficacy of intelligent networks for bearing fault diagnosis. Specifically, a set of Gabor wavelets are proposed to initialize the weight vectors of the SF to extract multi-scale and multi-directional fault features, and SE attention is used to guide the training of network weights. This approach aims to improve fault detection accuracy and computational efficiency. Two bearing datasets under different working conditions are employed to demonstrate the efficacy of the GWISF model. The experimental outcomes reveal that the proposed GWISF method outperforms other methods, exhibiting the highest diagnostic accuracy and robustness.

Accurate and efficient feature extraction remains a critical challenge in gearbox fault diagnosis, particularly under noisy operating conditions. To address this issue, a novel Gaborlet atlas-guided autoencoder network (GAAN) that leverages the unique feature extraction capabilities of the Gabor transform is proposed in this article. Despite its inherent inability to form an orthogonal basis, it possesses the capability to construct a compact frame under specific parametric conditions, thereby facilitating the utilization of the Gaborlet atlas in feature extraction tasks. Multiheaded attention mechanism is also used to assist training the weight matrix of the autoencoder network. The GAAN can capture multiscale and multidirectional fault features from noise signals, enhancing both interpretability and noise resilience of the diagnostic model. Extensive experiments on two noisy gearbox datasets demonstrate that GAAN outperforms traditional methods in both accuracy and computational efficiency, highlighting its potential for real-world industrial applications.