ML for machine perception, computational photography, computer vision, and video generation.

Work on Lytro Cinema (micro lens array light-field capture system) and Lytro Immerge (VR camera rig)

As part of the Computer Vision & Machine Learning group:

• Manage team of software engineers to develop key product features for light field processing and creative editing

• Research and develop algorithms for camera/sensor calibration and image processing for multi-view video systems (C++/SciPy)

• Solve computer vision problems in the areas of image segmentation, 3D scene reconstruction, depth estimation and sensor fusion

• Implement and optimize algorithms for a variety of platforms, runtime environments and performance requirements (C++, OpenCV, ceres-solver)

• Design quality metrics for machine learning networks and core CV algorithms

• Supervise entire process from research and evaluation to software design, build system and deployment.

Research and develop computer vision/computational photography algorithms for the Lytro Cinema camera.

• Research algorithms for image/video segmentation using probabilistic models

• Develop application-specific image processing and filtering algorithms in C++

• Maintain software for motion tracking and camera pose recovery

• Build data visualizations, prototype and debug algorithms using OpenCV/SciPy

• Author several patents from research activities

Research & develop core computer graphics technologies. Lead several projects with global teams.

• Implement algorithms to assemble, preview, render large hierarchies of 3D graphics data

• Achieve tight performance requirements via complexity analysis, caching & concurrent programming

• Negotiate and prioritize features internally and with 3rd party vendors

• Overview significant expansion of supported applications and number of users

Technologies: C++, Python, Intel TBB, boost test libraries

Develop tools to support Double Negative's expansion to Singapore and Vancouver in a fast-paced environment.

• Lead the development of a camera footage processing tool, ingesting 20+ digital formats and gigabytes of video data (daily!). Challenges included: rapidly evolving APIs, unreliable inputs, short deadlines, and large number of key actors in different time zones.

• Write a real time monitoring interface for globally distributed image conversion & data transfer queues (Python, QT, REST, SQL)

• Maintain the software release system for ~80 developers. Enable contributions by developers outside of the London office for the first time.