MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202641123834 A) filed by Vels Institute Of Science, Technology And Advanced Studies Vistas on October 15, 2026, for A System And A Method For Artificial Intelligence-Based Context-Aware Visual Content Filtering.
Inventors include S. Sumithra; and P. Sujatha.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: Disclosed is a system (100) for artificial intelligence-based context-aware detection, classification, and selective filtering of adult and sensitive visual content in digital images and videos. The system (100) comprises an input content acquisition unit (102), a pre-processing and normalization unit (104), a global visual-semantic analysis unit (106), a localized visual detection and region analysis unit (108), a visual evidence representation and fusion unit (110), a Large Language Model-based visual reasoning and content classification unit (112), a video frame evidence aggregation unit (114), a content moderation decision unit (116), a policy, threshold, and configuration unit (118), and a selective content filtering and enforcement unit (120). The system (100) integrates global semantic features and localized detection results into a structured visual evidence representation for contextual classification using Large Language Model-based reasoning. Based on classification results and configurable moderation policies, the system (100) selectively allows, blocks, blurs, or masks identified visual content. Fig. 1 is the reference figure.
Disclaimer: Curated by HT Syndication.
