Pallet racking is an essential element within warehouses, distribution centers, and manufacturing facilities. To guarantee its safe operation as well as stock protection and personnel safety, pallet racking requires continuous inspections and timely maintenance in the case of damage being discovered. Conventionally, a rack inspection is a manual quality inspection process completed by certified inspectors. The manual process results in operational down-time as well as inspection and certification costs and undiscovered damage due to human error. Inspired by the trend toward smart industrial operations, we present a computer vision-based autonomous rack inspection framework centered around YOLOv7 architecture. Additionally, we propose a domain variance modeling mechanism for addressing the issue of data scarcity through the generation of representative data samples. Our proposed framework achieved a mean average precision of 91.1%.

This paper presents a framework for the automated detection of Exudates, an early sign of Diabetic Retinopathy. The paper introduces a classification-extraction-superimposition (CES) mechanism for enabling the generation of representative exudate samples based on limited open-source samples. The paper demonstrates how the manipulation of Yolov5M output vector can be utilized for exudate extraction and super-imposition, segueing into the development of a custom CNN architecture focused on exudate classification in retinal based fundus images. The performance of the proposed architecture is compared with various state-of-the-art image classification architectures on a wide range of metrics, including the simulation of post deployment inference statistics. A self-label mechanism is presented, endorsing the high performance of the developed architecture, achieving 100% on the test dataset.

YOLOv7 algorithm have taken the object detection domain by the storm as its real-time object detection capabilities out ran all other previous algorithms both in accuracy and speed [1]. YOLOv7 advances the state of the art results in object detection by inferring more quickly and accurately than its contemporaries. In this paper, we are going to present our work of implementing this SOTA deep learning model on a soccer game play video to detect the players and football. As the result, it detected the players, football and their movement in real time. We also analyzed and compared the YOLOv7 results against its previous versions including YOLOv4, YOLOv5 and YOLO-R. The code is available at: https://github.com/RizwanMunawar/YOLO-RX57-FPS-Comparision

No abstract

Rice is a widely consumed food across the world. Whilst the world recovers from COVID-19, food manufacturers are looking to enhance their quality inspection processes for satisfying exportation requirements and providing safety assurance to their clients. Rice cultivation is a significant process, the yield of which can be significantly impacted in an adverse manner due to plant disease. Yet, a large portion of rice cultivation takes place in developing countries with less stringent quality inspection protocols due to various reasons including cost of labor. To address this, we propose the development of lightweight convolutional neural network architecture for the automated detection of rice leaf smut and rice leaf blight. In doing so, this research addresses the issue of data scarcity via a practical variance modeling mechanism (Domain Feature Mapping) and a custom filter development mechanism assisted through a reference protocol for filter suppression.