Meet Patel
About Meet Patel
Meet Patel is an AI Application Intern at Synaptics Incorporated, where he integrates Google's Gemma LLMs into the VS680 AI SoC. He has a strong background in machine learning and software engineering, with previous internships and notable research publications.
Work at Synaptics
Currently serving as an AI Application Intern at Synaptics Incorporated since 2023, Patel is based in San Jose, California. In this role, he led the integration of Google's Gemma LLMs into the Synaptics VS680 AI System on Chip (SoC) for real-time applications. His contributions have been recognized through his win at the Synaptics Tech Showcase 2023, where he presented an innovative edge AI Video Analytics demonstration.
Education and Expertise
Patel is pursuing a Master of Science in Computer Software Engineering at San Jose State University, expected to complete in 2024. He previously earned a Bachelor's degree in Computer Engineering from L.D. College of Engineering, graduating in 2021. His educational background provides a strong foundation in software engineering principles and machine learning techniques.
Background in Machine Learning
Before joining Synaptics, Patel gained valuable experience as a Machine Learning Research Intern at Ahmedabad University from 2020 to 2021. He also worked as a Software Engineer Intern at Royal Technosoft Pvt Ltd for four months in 2020. These roles allowed him to develop practical skills in machine learning and software development.
Research and Publications
Patel has contributed to the field of artificial intelligence through his published research titled 'Person Retrieval in Surveillance Videos Using Attribute Recognition' in the Springer journal of Ambient Intelligence & Humanized Computing in 2022. This work reflects his commitment to advancing AI technologies and their applications in real-world scenarios.
Technical Projects and Innovations
Patel has developed several technical projects, including a real-time YOLOv8 instance segmentation demo using GStreamer on the Yocto Linux platform. He optimized Vision Transformers with int8 quantization, achieving 95% accuracy while reducing processing time to 15ms per image. His work on enhancing real-time processing on Android Embedded OS resulted in an inference time reduction to 14ms per frame.