PREEN performs color style transfer via editor presets
Original title: PREEN: Exemplar-Based Color Style Transfer with Smartphone Editor Presets
PREEN predicts editor slider parameters from style references to achieve sub-second on-device latency.
PREEN casts color style transfer as regression onto 29 parameters of a deterministic image editor. Supervision uses a teacher model whose outputs are inverted into preset space. A CLIP-conditioned regressor is pretrained then refined with LoRA under CIELAB loss. On PST50 benchmark it leads baselines while exposing a continuous strength knob.
- Predicts preset vectors instead of images
- Sub-second on-device latency
- Leads baselines on fidelity metrics
- Interpretable slider values
- Single inference knob for vivid-to-default spectrum
From the original article
1 min read at huggingface.co
We present PREEN (Preset Regression via an Exemplar-based Editor Network), which casts color style transfer as regression onto the 29 parameters of a deterministic image editor. Given a content image and a style reference, PREEN predicts a preset, a vector of editor slider values that is interpretable and editor-consumable, not a raster image.
By Beomjin Ahn · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.