PyMAF
by HongwenZhang
[ICCV 2021, Oral] PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loop
AI summary
Human reconstruction
Reconstructs human 3D models from monocular images using a feedback loop to improve pose and shape estimation
- stars
- 600
- forks
- 67
- watching
- 13
Similar projects
Found by comparing what the projects do, not just their names.
Full-body model recovery
An open-source software project that enables the recovery of full-body 3D human models from monocular images.
Human reconstruction tool
This repository provides software tools and pre-trained models for estimating 3D human shape and pose from dense body parts
3D Pose Estimator
A PyTorch implementation of a 3D human pose estimation algorithm using weak supervision.
Pose estimator
A PyTorch implementation of a deep learning model for 3D multi-person pose estimation from a single RGB image
Dense reconstruction model
An implementation of CloudWalk's DenseBody paper using PyTorch
Human Pose Data
Provides pseudo-GT 3D human pose data and pre-trained models for training 3D pose estimation algorithms
Human model toolkit
Provides a modular framework and tools for working with 3D human parametric models in computer vision and graphics
Human pose and shape estimator
An implementation of an expressive human pose and shape estimation system using deep learning techniques
Pose estimator
A PyTorch toolkit for 2D human pose estimation with support for multi-thread data loading, multi-GPU training, and visualization.
Image-to-3D capture framework
A software framework for capturing 3D human body and facial features from single images using machine learning models.
Action Recognition Model
Replication of a PyTorch model for action recognition and detection from skeleton data
Scene reconstruction
An implementation of a 3D scene reconstruction method from a single image using computer vision and machine learning techniques
Human generator
A system for generating 3D human models from 2D image collections using deep learning
Image-based 3D modeling
Develops a method to reconstruct 3D wireframe models from a single image using deep learning
Model alignment
This project demonstrates the effectiveness of reinforcement learning from human feedback (RLHF) in improving small language models like GPT-2.
