MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641076516 A) filed by Cvr College Of Engineering on June 17, 2026, for Intelligent System For Early Autism Spectrum Disorder Detection Using Vision Transformer-Based Facial Image Analysis.
Inventors include Shaik Janbhasha; Moghal Irfan Pasha; Nallametti Nagarjuna; and Moghal Yaseen Pasha.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder, the early diagnosis of which, provide timely intervention and better developmental outcomes, most of the existing methods relying on machine learning and CNN-based methods, which suffers from capturing long-range spatial dependency and complex relationship among facial regions. The present invention provides an intelligent system to detect ASD early on the basis of facial image analysis by Vision Transformer. Facial images are taken from publicly available facial image datasets. The proposed method can effectively model fine-grained local facial features and global spatial relationship between the whole face. The invention uses a self-attention mechanism to learn highly discriminative representation of facial images associated with ASD and to be able to classify it accurately and automatically. The system proposed is a non-invasive, scalable, robust and efficient solution for large-scale screening and clinical decision support applications for ASD.
Disclaimer: Curated by HT Syndication.
