Publication Type : Conference Proceedings
Publisher : IEEE
Source : 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT)
Url : https://ieeexplore.ieee.org/abstract/document/10724740
Campus : Bengaluru
School : School of Engineering
Year : 2024
Abstract : In this era of rapid technological advancement, the detection of metal damage has become increasingly crucial across various industries. The proposed work “Cutting-Edge Steel Damage Classification Powered by Cloud Technology” aims to build an advanced system for real-time steel damage classification utilizing AWS services includes Rekognition, Lambda, and API Gateway that enables users to upload multiple images simultaneously for damage detection and classification, which will be an innovative approach to industrial defects detection which leads to their prevention. Such benefits of fast, accurate and scalable defect detection make it a critical asset for businesses expecting for exact and real-time solutions in every aspect. This study uses F1 score, precision, recall, and assumed threshold are as evaluation parameters. The use of AWS services is not just helping with model training and deployment, but it also provides failover mechanism and scaling capabilities to accommodate the varying workload demands. The model has the advantage of having rapid training time: 1.005 hours this shows the cloud infrastructure scalability and computational efficiency in dealing with the tasks requiring machine learning of complex nature.
Cite this Research Publication : Pennabadi Devendra Reddy, K Satya Sampath Reddy, P Gnaneswarachary, P Hemanth Reddy, Mrudula Prashanth, Shinu M. Rajagopal, Cutting-Edge Steel Damage Classification Powered by Cloud Technology, 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT), 2024