MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202641107010 A) filed by Easwari Engineering College; and Srm Institute Of Science And Technology, Ramapuram Campus on September 05, 2026, for Open-Set Anomaly Detection System And Method For Identifying Unusual Exoplanet Atmospheric Chemistry From Observed Transit Spectra.
Inventors include Shreenidhi B L; B Neha; and Mrs. P Ramya.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: Automated identification of exoplanet atmospheric chemistry from transmission spectra is increasingly urgent given the volume of new targets from JWST and other space telescopes since 2022. Closed-set classifiers cannot discover novel compositions, as they only assign spectra to predefined classes. We propose an open-set anomaly detection framework that learns a model of normal atmospheric composition directly from transmission spectra using self-supervised learning, then flags outliers via standard baselines. A hybrid anomaly score combines reconstruction error with latent-space distance, while an occlusion-based explainability module attributes scores to wavelength regions and candidate molecular species. Unlike prior work limited to the synthetic Ariel Big Challenge (ABC) database, this pipeline supports both synthetic and observational spectra, including data from the Mikulski Archive for Space Telescopes (MAST). Combining self-supervised learning with interpretable anomaly attribution, this work offers a scalable triage tool for future missions like Ariel, aiding discovery of uncharacterized exoplanet chemistries. Total Number of Words in Abstract: 147
Disclaimer: Curated by HT Syndication.
