Swift Server Side Presentation
A talk covering practical server-side Swift work and engineering lessons.
Software Engineer • Imaging Systems • Computer Vision
Building robust visual products with a research mindset.
I am Mathieu Barnachon, a software engineer with a background in computer vision, motion capture, imaging pipelines, and applied machine learning. My work spans production systems, cloud tooling, and research-driven product development.
- 2016–Today Product and platform engineering at StyleShoots
- 2013–2016 Camera tracking and AR systems at Ncam
- Ph.D. 2013 Computer Science, Université Lyon 1
About
Applied engineering with strong foundations in vision research.
Since March 2016, I have been working at StyleShoots, where I helped design the video composition pipeline used in the company’s integrated studio systems. I also contributed to the RAW image processing pipeline and the color styling workflows applied to the resulting media.
More recently, my work has focused on cloud products and integrated tooling, including custom event tracking and image and video upload workflows. Alongside product engineering, I have explored proof-of-concept work in machine learning, including deep learning, K-Means, and SVM-based approaches.
Before StyleShoots, I worked at Ncam Technologies on camera tracking and augmented reality systems for broadcast and film production. Earlier academic work included a Ph.D. in Computer Science from Université Lyon 1, postdoctoral research in Auckland, and research collaborations in Canada.
Selected Work
Projects, demos, and product moments.
A mix of production work, talks, demos, and earlier research output.
Live Machine Walkthrough
An overview of an integrated capture workflow in production.
Eclipse Machine Introduction
A product introduction focused on studio automation and workflow design.
Live Machine Introduction
A compact introduction to a connected production system for image and video.
Ncam Showreel
Examples of camera tracking and augmented reality capabilities in action.
Human Action Recognition
A research demo around streamed motion capture and online action recognition.
Experience
Career highlights.
2016–Present
StyleShoots
Designed core media-processing systems, including video composition and RAW image workflows, then moved into cloud-connected tools and platform services.
- Video composition pipeline for integrated studio systems
- RAW image processing and color styling workflows
- Cloud event tracking and upload tooling
2013–2016
Ncam Technologies
Worked on camera tracking and real-time augmented reality systems for broadcast and film environments, with a strong emphasis on accuracy and speed.
- Tracking for film and broadcast use cases
- Real-time AR in unconstrained environments
- Near real-time film-quality tracking
2013
Postdoctoral Research, University of Auckland
Research on action recognition in office environments and real-time stereo vision improvements.
2012
Research Stay, University of Windsor
Research collaboration focused on one-shot learning in pattern recognition.
2009–2013
Ph.D., Université Lyon 1
Doctoral work on human action recognition from motion capture data.
Contact
Open to technical conversations and collaborations.
The easiest way to reach me is by email. You can also find my work and profile on GitHub and LinkedIn.