imagesam3dbodymapper

February 4, 2026 · View on GitHub

SAM 3D Body (3DB) is a promptable model for single-image full-body 3D human mesh recovery (HMR).

SAM 3D Body(3DB)是一种可提示的模型,用于从单张图像中恢复完整人体的3D网格(HMR)。

Type 算子类型: mapper

Tags 标签: gpu, image

🔧 Parameter Configuration 参数配置

name 参数名type 类型default 默认值desc 说明
checkpoint_path<class 'str'>''Path to SAM 3D Body model checkpoint.
detector_name<class 'str'>'vitdet'Human detection model for demo (Default vitdet, add your favorite detector if needed).
segmentor_name<class 'str'>'sam2'Human segmentation model for demo (Default sam2, add your favorite segmentor if needed).
fov_name<class 'str'>'moge2'FOV estimation model for demo (Default moge2, add your favorite fov estimator if needed).
mhr_path<class 'str'>''Path to MoHR/assets folder (or set SAM3D_mhr_path).
detector_path<class 'str'>''Path to human detection model folder (or set SAM3D_DETECTOR_PATH).
segmentor_path<class 'str'>''Path to human segmentation model folder (or set SAM3D_SEGMENTOR_PATH).
fov_path<class 'str'>''Path to fov estimation model folder (or set SAM3D_FOV_PATH).
bbox_thresh<class 'float'>0.8Bounding box detection threshold.
use_mask<class 'bool'>FalseUse mask-conditioned prediction (segmentation mask is automatically generated from bbox).
visualization_dir<class 'str'>NoneDirectory to save visualization results. If None, no visualization will be saved.
tag_field_name<class 'str'>'sam_3d_body_data'Field name for storing the results.
args''
kwargs''