metadata

language: en
license: cc-by-nc-4.0
tags:
  - sapiens

Seg-Sapiens-1B

Model Details

Sapiens is a family of vision transformers pretrained on 300 million human images at 1024 x 1024 image resolution. The pretrained models, when finetuned for human-centric vision tasks, generalize to in-the-wild conditions. Sapiens-1B natively support 1K high-resolution inference. The resulting models exhibit remarkable generalization to in-the-wild data, even when labeled data is scarce or entirely synthetic.

Developed by: Meta
Model type: Vision Transformer
License: Creative Commons Attribution-NonCommercial 4.0
Task: seg
Format: original
File: sapiens_1b_goliath_best_goliath_mIoU_7994_epoch_151.pth

Model Card

Image Size: 1024 x 768 (H x W)
Num Parameters: 1.169 B
FLOPs: 4.647 TFLOPs
Patch Size: 16 x 16
Embedding Dimensions: 1536
Num Layers: 40
Num Heads: 24
Feedforward Channels: 6144

More Resources

Repository: https://github.com/facebookresearch/sapiens
Paper: https://arxiv.org/abs/2408.12569
Demo: https://huggingface.co/spaces/facebook/sapiens-seg
Project Page: https://about.meta.com/realitylabs/codecavatars/sapiens
Additional Results: https://rawalkhirodkar.github.io/sapiens
HuggingFace Collection: https://huggingface.co/collections/facebook/sapiens-66d22047daa6402d565cb2fc

Uses

Seg 1B model can be used to perform 28 class body part segmentation on human images.