Headline

Computer Vision · Deep Learning · Face Recognition · VLMs | Research Engineer @ Mercedes-Benz R&D | MS by Research, IIIT-H | IROS & NeurIPS

About

I build computer vision systems that have to work in the real world — inside moving vehicles, under hard latency and robustness constraints, where a model either holds up or it doesn't. At Mercedes-Benz R&D, I work on the in-cabin sensing algorithms behind face recognition, occupant analysis, and related perception tasks, owning problems from proof-of-concept through to production. Earlier, during my MS by Research at IIIT Hyderabad's CVIT lab, my work on open-world object detection for road scenes was published at IROS and a NeurIPS workshop. The through-line is turning research into things that actually ship. I spend most of my attention on the unglamorous parts: data quality, bridging RGB/NIR modality gaps, knowledge distillation, and getting models small and reliable enough to deploy. Outside work, I fine-tune vision-language models for practical use cases and build ML tooling for problems I find interesting.

Location

Bengaluru

AI stack

claude code

claude

github copilot

hugging face

opencode

Weights & Biases Weave

Work experience

Computer Vision and Deep Learning Engineer at Mercedes-Benz Research and Development India, Apr 2022 - Present · 4 yrs 6 mos, Bengaluru, Karnataka, India, - Build production computer vision systems for in-cabin automotive sensing — face recognition, depth estimation, and occupant analysis. - Led an in-cabin face recognition system end to end, from data strategy through deployment, to meet strict accuracy and robustness targets. - Designed a dual-level authentication approach balancing low-latency personalization with high-assurance security for sensitive use cases. - Developed cross-domain (RGB/NIR) face recognition and IR-to-RGB colorization methods that reduced reliance on large-scale data collection. - Built software-driven occupant analysis models, including IR-based weight estimation, replacing hardware sensors. Work across software, quantization, validation, and platform teams; contributor to filed patents.

Graduate Research Scholar at Center for Visual Information Technology, IIIT Hyderabad, Jan 2019 - Mar 2022 · 3 yrs 3 mos, Hyderabad, Telangana, India

Software Engineer at Celstream Technologies, Aug 2014 - Sep 2016 · 2 yrs 2 mos, Bangalore Urban, Karnataka, India

Education

Master's degree, Computer Science, International Institute of Information Technology Hyderabad (IIITH)

Bachelor's degree, Computer Science, Visvesvaraya Technological University

Public links

github: https://github.com/deepaksinghcv

deepaksinghcv.github.io: https://deepaksinghcv.github.io/