MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641075867 A) filed by Cvr College Of Engineering on June 19, 2026, for Artificial Intelligence-Based Skin Lesion Classification System Utilizing Data Augmentation And Class Balancing Techniques..
Inventors include Gera Vijaya Nirmala; and Dr K. A. Jyotsna.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: The system and procedure for detecting skin cancer utilizing adaptive deep learning algorithms is the subject of the current invention. The invention uses a Convolutional Neural Network combined with data augmentation, class imbalance control, and Adaptive Moment Estimation (Adam) optimization. Prior to feature extraction and classification, medical imaging scans are obtained and pre-processed. To overcome small training datasets, data augmentation techniques like rotation, flipping, translation, scaling, shearing, and color jittering are used. Balancing strategies like oversampling, under sampling, class weighting, focus loss, and Synthetic Minority Oversampling Technique (SMOTE) are used to mitigate class imbalance. The combined framework increases model generalization, decreases over fitting, increases classification accuracy, and offers dependable automated skin cancer diagnosis. Applications for computer-aided dermatological screening, cloud-based healthcare platforms, and medical diagnostic systems can all benefit from the concept.
Disclaimer: Curated by HT Syndication.
