ARKitScenes
Indoor scene dataset
A large dataset of 3D indoor scenes captured with RGB-D data and processed for use in machine learning applications
This repo accompanies the research paper, ARKitScenes - A Diverse Real-World Dataset for 3D Indoor Scene Understanding Using Mobile RGB-D Data and contains the data, scripts to visualize and process assets, and training code described in our paper.
674 stars
26 watching
59 forks
Language: Python
last commit: almost 2 years agoRelated projects:
| Repository | Description | Stars |
|---|---|---|
| A large dataset of 3D object and room layout annotations on RGB videos, designed to test automatic scene understanding methods. | 106 | |
| A large-scale dataset of indoor visual data with 360-degree panoramas and annotated room layouts for computer vision applications | 178 | |
| ARKit project providing an example of building augmented reality experiences with SceneKit and ARCore-like functionality | 374 | |
| A synthetic dataset designed to improve performance in indoor scene perception tasks by providing detailed labels and photorealistic images. | 40 | |
| Provides pre-trained deep learning models for scene classification on the Places365 dataset | 1,935 | |
| An image classification model for recognizing physical places and locations | 41 | |
| An ARKit sample application demonstrating various augmented reality features and user interactions. | 77 | |
| Demonstrates how to track and occlude objects with real-world geometry using ARKit+SceneKit in iOS apps. | 349 | |
| A collection of ARKit samples providing various practical examples and techniques for building augmented reality experiences. | 545 | |
| This project demonstrates ARKit-based motion of the solar system | 617 | |
| A PyTorch implementation of a deep learning-based method for generating interactive scenes with specified object attributes and relations | 188 | |
| A dataset of 2D images and 3D data generated from the Grand Theft Auto game engine for object localization research. | 135 | |
| A 3D scene graph library with a component-based node system and realistic rendering pipeline for WebGL | 369 | |
| Specification for delivering and persisting large 3D geographic data in a scalable and format-agnostic way | 321 | |
| Tools and scripts to visualize large datasets from the KITTI dataset in Blender using Python | 72 |