Skip to content
Meet Pandya

Hello,
I’m Meet.

I’m a computer vision engineer. I turn images into useful measurements and build the systems that bring those models into production.

I build computer vision systems for under-vehicle inspection.

Technical lead for DepScan
at Vehant Technologies.

Explore my work

Selected work

Faster vehicle scans

30% lower mosaic latency

Optimized shift detection and image mosaicing for sharper under-vehicle scans.

More about this work
Baseline100
Optimized70
Relative generation latency · baseline = 100
Read the DepScan paper

Learning from less data

90% accuracy · 10 samples per category

Developed an incremental vehicle classifier for new categories using 10 samples per category.

Fewer false alarms

25% fewer false positives

Reduced false alarms in under-vehicle inspection through one-class anomaly detection.

Bringing models into production

Framework-agnostic model deployment

Built a framework-agnostic Python inference server and improved model release workflows.

Selected publications

DepScan: 3D Under Vehicle Scanning SystemDepScan: 3D under-vehicle scanning2023 · AIML SystemsSegregating and Recognizing Human Actions from Video Footages Using LRCN TechniqueHuman action recognition using LRCN2020 · AMLTA

A little more about me

My work spans image mosaicing, few-shot recognition, anomaly detection, and stereo vision. I work in Python, C++, and CUDA, from the image-processing pipeline to the inference server.

Education & teaching

I earned my M.Tech. in Software Systems at DA-IICT, where I also worked as a teaching assistant in Computer Architecture and Linux Systems.

Before that, I completed my B.Tech. in Computer Engineering at Ganpat University.