Open-source Sapiens-seg-0.3b-torchscript model - Supports 1K high-resolution inference with excellent generalization ability

Sapiens Seg 0.3b Torchscript

Developed by facebook

Sapiens is a family of vision Transformer models pre-trained on 300 million 1024 x 1024 resolution human images, supporting 1K high-resolution inference, demonstrating exceptional generalization to real-world data even with scarce or entirely synthetic labeled data.

Image Segmentation English#High-resolution human segmentation #28-class body part recognition #300 million image pre-training

Downloads 56

Release Time : 9/9/2024

Model Overview

The Seg 0.3B model can be used for 28-class body part segmentation of human images.

Model Features

High-resolution support

Native support for 1K high-resolution inference with an image size of 1024 x 768.

Strong generalization capability

Demonstrates exceptional generalization to real-world data even with scarce or entirely synthetic labeled data.

Efficient computation

Computational load of 1.242 TFLOPs, patch size of 16 x 16, and embedding dimension of 1024.

Model Capabilities

Human image segmentation

High-resolution image processing

28-class body part recognition

Use Cases

Medical imaging

Body part segmentation

Used for precise segmentation of human body parts in medical imaging.

Supports 28-class body part segmentation.

Virtual reality

Virtual avatar creation

Used to create high-precision human models in virtual reality applications.

High-resolution support enhances the detail representation of virtual avatars.

🚀 Seg-Sapiens-0.3B-Torchscript

Seg-Sapiens-0.3B-Torchscript is a vision transformer model for human image segmentation, pretrained on a large - scale human image dataset and capable of high - resolution inference.

🚀 Quick Start

This model is designed for human - centric vision tasks, especially image segmentation. It can be used to perform 28 - class body part segmentation on human images.

✨ Features

Sapiens is a family of vision transformers pretrained on 300 million human images at 1024 x 1024 image resolution. When finetuned for human - centric vision tasks, it generalizes well to in - the - wild conditions.
Sapiens - 0.3B natively supports 1K high - resolution inference. The models show remarkable generalization to in - the - wild data, even with scarce labeled data or entirely synthetic data.

📦 Installation

No installation steps are provided in the original README, so this section is skipped.

💻 Usage Examples

No code examples are provided in the original README, so this section is skipped.

📚 Documentation

Model Details

Developed by: Meta
Model type: Vision Transformer
License: Creative Commons Attribution - NonCommercial 4.0
Task: seg
Format: torchscript
File: sapiens_0.3b_goliath_best_goliath_mIoU_7673_epoch_194_torchscript.pt2

Model Card

Property	Details
Image Size	1024 x 768 (H x W)
Num Parameters	0.336 B
FLOPs	1.242 TFLOPs
Patch Size	16 x 16
Embedding Dimensions	1024
Num Layers	24
Num Heads	16
Feedforward Channels	4096

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

🔧 Technical Details

No detailed technical implementation information is provided in the original README, so this section is skipped.

📄 License

The model is licensed under the Creative Commons Attribution - NonCommercial 4.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご