Open-source alternatives to Sergeyprokudin Bps

Grounded-Segment-Anything is a suite of specialized tools for multimodal visual analysis, text-based segmentation, and generative image editing. It integrates text-to-bounding-box detection and high-precision image segmentation masks to function as a text-based image segmenter and an automated visual labeling tool. The project enables text-driven image editing by identifying objects through natural language to perform inpainting and element replacement. It further extends visual analysis into three dimensions, allowing for 3D human reconstruction and the generation of 3D bounding boxes from t

paper project page

BodyNet: Volumetric Inference of 3D Human Body Shapes, ECCV 2018

This repository is the official implementation of paper: "Unsupervised Point Cloud Pre-training via Occlusion Completion"

This repository contains the code for the following paper:

Created by Itai Lang, Asaf Manor, and Shai Avidan from Tel Aviv University.

Paper - Supplementaty - Project Website - Arxiv - Video - Published in CVPR 2020.

This repository contains the source code to support the paper: Shape Adaptor: A Learnable Resizing Module, by Shikun Liu and Adobe Research Team.

Project Page Paper Code

Exchangeable Neural ODE for Set Modeling

Yiqun Lin, Zizheng Yan, Haibin Huang, Dong Du, Ligang Liu, Shuguang Cui, Xiaoguang Han, "FPConv: Learning Local Flattening for Point Convolution", CVPR 2020 [paper](https://arxiv.org/abs/2002.10701)

Authors: Francesco Milano, Antonio Loquercio, Antoni Rosinol, Davide Scaramuzza, Luca Carlone

VIBE is a 3D human pose estimation framework designed to reconstruct human body shapes and poses from video frames. It functions as a toolkit for predicting parameters of the SMPL human body model to generate 3D mesh sequences. The system includes a 3D motion data exporter to convert predicted pose sequences into standard 3D file formats for use in graphics and animation software. It also provides a structured training pipeline for preparing datasets and training models to estimate body shapes from images. Its capabilities cover computer vision for estimating body pose and shape, as well as

We consider neural networks with rational activation functions. The choice of the nonlinear activation function in deep learning architectures is crucial and heavily impacts the performance of a neural network. We establish optimal bounds in terms of network complexity and prove that rational…

AMASS is a large database of human motion unifying different optical marker-based motion capture datasets by representing them within a common framework and parameterization. AMASS is readily useful for animation, visualization, and generating training data for deep learning.

NVAE is a deep hierarchical variational autoencoder that enables training SOTA likelihood-based generative models on several image datasets.

Coder for "On the Continuity of Rotation Representations"

Created by Pedro Hermosilla , Tobias Ritschel , Timo Ropinski .

MeshCNN is a general-purpose deep neural network for 3D triangular meshes, which can be used for tasks such as 3D shape classification or segmentation. This framework includes convolution, pooling and unpooling layers which are applied directly on the mesh edges.

Platform for designing and evaluating Graph Neural Networks (GNN)

This repository is the official PyTorch implementation of "Implicit Graph Neural Networks".

Created by Jonas Schult, Francis Engelmann, Theodora Kontogianni and Bastian Leibe from RWTH Aachen University.

Vincent Sitzmann\, Julien N. P. Martel\, Alexander W. Bergman, David B. Lindell, Gordon Wetzstein Stanford University, \*denotes equal contribution

This repository is the official PyTorch implementation of IF-Defense: 3D Adversarial Point Cloud Defense via Implicit Function based Restoration, where we study adversarial defense in 3D point cloud and propose an deep implicit function based defense method that achieves state-of-the-art (SOTA)…

This is the project webpage of our CVPR 2020 work. RPM-Net is a deep-learning approach designed for performing rigid partial-partial point cloud registration for objects in a iterative fashion. Our paper can be found on Arxiv (supplementary).

This is the official implementation for paper "Patch-base progressive 3D Point Set Upsampling".