FeyNoBg: A SOTA Model For Background Removal

gist

Feyn presents FeyNoBg, a 263-million-parameter background-removal model that leads four of eight reported benchmarks and comes within two percent of the leader on the rest. The work attributes the gain to modestly expanding BiRefNet while preserving compatible pretrained weights, then training on a deliberately capped and diverse mixture of segmentation and matting datasets. Feyn also releases NoBg, the common inference and training library used for the experiments.

ideas

• Treat recognition and boundary precision as coupled skills. Good cutouts require both identifying the whole subject and tracing translucent or fine edges, so optimizing either skill alone can degrade the other.
• Add capacity without discarding prior learning. FeyNoBg expands BiRefNet’s most information-rich feature stage from 18 to 24 blocks while retaining every compatible pretrained weight.
• Balance the data mixture deliberately. Per-source caps prevent large datasets from dominating, while varied scenes teach the model to handle camouflage, crowds, portraits, high resolution, and fine structures.
• Normalize unlike annotations to one target. Converting masks and alpha mattes into binary foreground masks creates consistent supervision, though it gives up soft-opacity training signals.
• Package the experimental substrate. NoBg exposes a common interface for running, comparing, and fine-tuning background-removal models so follow-on work needs fewer bespoke adapters.

quotes