MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611089856 A) filed by Shantanu Pandya on July 23, 2026, for A System And Method For Resource-Optimized Multi-Modal Ai-Driven Programmatic Video Editing And Rendering.
Inventor includes Shantanu Pandya.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: A computer-implemented system and method for resource-optimized, multi-modal AI-driven programmatic video editing is disclosed. A local application server (101) uploads a target video file and a natural-language instruction to an external multi-modal large language model server (102), which returns a structured JSON payload defining video and audio operations. An editing orchestration engine (105) executes the operations. A Digital Signal Processing audio engine (106) performs stationary noise reduction on the audio track in sequential, time-sliced chunks to limit memory allocation. A fallback graphics-based text-rendering module (107) constructs stylized caption overlays using a raster graphics library when a system-level rendering dependency is unavailable. The composite output video is rendered using thread- constrained, memory-efficient encoding, solving system-dependency failures and memory overheads present in conventional programmatic video editing tools.
Disclaimer: Curated by HT Syndication.
