facebook/sapiens-normal-0.6b-bfloat16

Normal-Sapiens-0.6B-Bfloat16

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-0.6B 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: normal
• —Format: bfloat16
• —File: sapiens0.6bnormalrenderpeopleepoch200_bfloat16.pt2

Model Card

• —Image Size: 1024 x 768 (H x W)
• —Num Parameters: 0.664 B
• —FLOPs: 2.583 TFLOPs
• —Patch Size: 16 x 16
• —Embedding Dimensions: 1280
• —Num Layers: 32
• —Num Heads: 16
• —Feedforward Channels: 5120

More Resources

• —Repository: https://github.com/facebookresearch/sapiens
• —Paper: https://arxiv.org/abs/2408.12569
• —Demo: https://huggingface.co/spaces/facebook/sapiens-normal
• —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

Normal 0.6B model can be used to estimate surface normal (XYZ) on human images.