nikopueringer/CorridorKey

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

14594 stars · 891 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

CorridorKey is a Python-based neural network tool that unmixes green/blue screen footage, reconstructing true foreground color and linear alpha per pixel for realistic composites. It preserves semi-transparent details like hair, motion blur, and translucency that traditional keyers and binary AI roto tools destroy.

Use cases

• key green screen footage with soft edges
• extract alpha matte from chroma key plates
• remove green spill from composited subjects
• preserve motion blur and hair in a key
• unmix blue screen plates for VFX compositing
• replace manual rotoscoping of green screen shots

When to choose

• you need physically accurate straight color and linear alpha output
• traditional keyers fail on semi-transparent or motion-blurred edges
• you want to avoid manual rotoscoping and edge matte work
• you work with 4K green or blue screen plates

When to avoid

• you need a battle-tested production tool - this is a brand new release
• you need blue screen support on Apple MLX backend (green only until the checkpoint ships)
• you want a GUI - use the EZ-CorridorKey companion instead
• you need real-time keying during capture

Facets

application · maturity experimental

image-processing machine-learning deep-learning image-processing media computer-vision python cross-platform cli green-screen chroma-key vfx compositing alpha-matte neural-network despill roto video gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="nikopueringer/CorridorKey")