Optimizing Portrait Lighting at Capture-Time Using a 360 Camera as a Light Probe
summary
A capture-time tool uses a 360 camera as a light probe and HDR environment maps plus a generic 3D head to compute how to orient a subject to match studio lighting. It uses precomputed radiance transfer for fast rendering on mobile, scores orientations by a light/dark weighting function, and finds specific optima (example: 23 degrees optimal rotation). The prototype suggests external phone-screen placement as a secondary source, reduces user effort, and produces demonstrably better portraits.
AI-written summary · about
chapters (8)
- 0:00Start
- 0:20Lighting Style
- 0:56Badger
- 2:59Capture the Environment Map
- 4:18Reorientation Guidance
- 10:56Left Rim
- 11:53The Target Lighting Style
- 15:44Lighting Styles
related talks
06mMaking Portrait mode yourself
50mPhotogrammetry workflow and the tech behind the de-lighting tool
26mCameraX: Portrait of an API
20mInteractive Room Capture on 3D-Aware Mobile Devices
48mColour Management – A Bird’s Eye View for Digital Artists
18mCamTrackPoint: Camera-Based Pointing Stick Using Transmitted Light through Finger
41mComputer Vision for Remote AR
32mSelf-Portraits in JavaScript