I will build high performance computer vision models for image and video analysis

As a published AI researcher (WACV, ECCV), I deeply understand the mathematics and architecture behind modern vision systems. I have successfully engineered complex, multimodal pipelines that process real-time video on edge devices with sub-200ms latency. If your project requires cutting edge precisionlike zero-shot object detection, temporal video segmentation, or automated visual QAyou are in the right place.

What I Can Build For You:

• Advanced Object Detection & Tracking: Implementing fast, highly accurate pipelines for real-time tracking and OCR.
• Video Analytics & Activity Recognition: Complex temporal segmentation and human activity understanding using lightweight 2D skeleton sequences.
• Multimodal VLM Systems: Fusing Vision-Language Models for complex image/video Question & Answering and automated reporting.
• Edge AI Optimization: Compressing and distilling massive teacher models into compact, ultra-fast student models for mobile or low-compute environments.

Why Choose Me?

• Published Expert: Author of novel self-supervised video alignment frameworks.
• Production-Ready Efficiency: Proven record of slashing model latency from 20 seconds to 1-2 seconds without sacrificing accuracy