deepomf's figure segmentation dataset

jerryli27 / AniSeg

A faster-rcnn model for anime character segmentation.

Apache License 2.0

202 stars 15 forks source link

deepomf's figure segmentation dataset #9

Open gwern opened 4 years ago

gwern commented 4 years ago

If you're interested in doing further training, deeppomf has an image dataset scraped from yiff.party (includes NSFW images) where PSDs are used to generate pixel-perfect segmentations of figures: https://drive.google.com/file/d/1vK7R_FiD3_pC892sWTy_fOA_HF6E2SKx/view

Artists provide PSDs in layers, where 1 layer is the figure and the rest the background; so pixel-perfect-by-construction segmentation masks can be automatically created for training.

ghost commented 4 years ago

@gwern Awesome! Do you have pretrained model on this dataset or scripts to disassemble this PSD files?

gwern commented 4 years ago

No, I've never used it. Just deepomf mentioned it as a way to improve on the existing segmentation without extremely laborious hand-annotation or using weak supervision from a classifier/tagger like DeepDanbooru, so I thought I'd mention that this dataset (and approach in general) exists.