Nishant Patil

Title: Nishant Patil and Machine-Learning-Based Video Encoding.

Introduction

Nishant Patil is an inventor associated with Mountain View, California, and Gdm Holding LLC. His patent work addresses video encoding, applying machine learning and feedback control to help select quantization parameters for video frames.

Latest Patents

Patil is named on 1 patent: “Rate Control Machine Learning Models with Feedback Control for Video Encoding.” The invention describes methods and systems for encoding a sequence of video frames. For each frame, a machine-learning model processes a feature embedding and scores possible quantization parameter values. The system selects a value, tracks the cumulative data needed for the current and preceding frames, and uses a feedback-control criterion to determine whether to update the selection before encoding the frame.

Career Highlights

The patent reflects work at the intersection of machine learning, feedback control, and video compression. Its approach uses frame-level information and cumulative data measurements to guide encoding decisions.

Collaborations

Chenjie Gu and Hongzi Mao are identified as Patil’s coworkers. They are also associated with the patent’s collaborative technical context.

Conclusion

Nishant Patil’s patent contribution focuses on using machine learning and feedback control to improve quantization decisions in video encoding.