Maayan Yedidia

Patent application describing systems and methods for matching textless elements to texted elements in video content to support localization and automated replacement of texted material.

Patent application describing methods for generating summary prompts that combine audio and visual insights (transcripts, speaker IDs, OCR, object/face recognition) and using those prompts with LLMs to create multimedia content summaries.

Product announcement / blog describing enhancements to Azure AI Video Indexer for multi-modal video summarization using generative models.

Patent family covering multichannel audio speech classification techniques, including transcoding audio channels, computing per-channel power/correlation values and classifying speech-based communications; published/issued across jurisdictions.