The timely diagnosis and identification of apple leaf diseases is essential to prevent the spread of diseases and ensure the sound development of the apple industry. Convolutional neural networks (CNNs) have achieved phenomenal success in the area of leaf disease detection, which can greatly benefit the agriculture industry. However, their large size and intricate design continue to pose a challenge when it comes to deploying these models on lightweight devices. Although several successful models (e.g., EfficientNets and MobileNets) have been designed to adapt to resource-constrained devices, these models have not been able to achieve significant results in leaf disease detection tasks and leave a performance gap behind. This research gap has motivated us to develop an apple leaf disease detection model that can not only be deployed on lightweight devices but also outperform existing models. In this work, we propose AppViT, a hybrid vision model, combining the features of convolution blocks and multi-head self-attention, to compete with the best-performing models. Specifically, we begin by introducing the convolution blocks that narrow down the size of the feature maps and help the model encode local features progressively. Then, we stack ViT blocks in combination with convolution blocks, allowing the network to capture non-local dependencies and spatial patterns. Embodied with these designs and a hierarchical structure, AppViT demonstrates excellent performance in apple leaf disease detection tasks. Specifically, it achieves 96.38% precision on Plant Pathology 2021—FGVC8 with about 1.3 million parameters, which is 11.3% and 4.3% more accurate than ResNet-50 and EfficientNet-B3. The precision, recall and F score of our proposed model on Plant Pathology 2021—FGVC8 are 0.967, 0.959, and 0.963 respectively.

Augmentation of abnormal cells in the brain causes brain tumor (BT), and early screening and treatment will reduce its harshness in patients. BT’s clinical level screening is usually performed with Magnetic Resonance Imaging (MRI) due to its multi-modality nature. The overall aims of the study is to introduce, test and verify an advanced image processing technique with algorithms to automatically extract tumour sections from brain MRI scans, facilitating improved accuracy. The research intends to devise a reliable framework for detecting the BT region in the two-dimensional (2D) MRI slice, and identifying its class with improved accuracy. The methodology for the devised framework comprises the phases of: (i) Collection and resizing of images, (ii) Implementation and Segmentation of Convolutional Neural Network (CNN), (iii) Deep feature extraction, (iv) Handcrafted feature extraction, (v) Moth-Flame-Algorithm (MFA) supported feature reduction, and (vi) Performance evaluation. This study utilized clinical-grade brain MRI of BRATS and TCIA datasets for the investigation. This framework segments detected the glioma (low/high grade) and glioblastoma class BT. This work helped to get a segmentation accuracy of over 98% with VGG-UNet and a classification accuracy of over 98% with the VGG16 scheme. This study has confirmed that the implemented framework is very efficient in detecting the BT in MRI slices with/without the skull section.