kubric

Video generator

Automates video data generation with realistic physics simulation and annotations for machine learning training

A data generation pipeline for creating semi-realistic synthetic multi-object videos with rich annotations such as instance segmentation masks, depth maps, and optical flow.

GitHub

2k stars
40 watching
235 forks
Language: Jupyter Notebook
last commit: almost 2 years ago
Linked from 1 awesome list


Backlinks from these awesome lists:

Related projects:

RepositoryDescriptionStars
kubeflow/pipelinesA tool for building and managing machine learning workflows on Kubernetes.3,634
kubeflow/kubeflowAn ecosystem of Kubernetes-based components for each stage in the AI/ML lifecycle14,472
google-research-datasets/objectronA dataset of short video clips with 3D bounding box annotations for object detection and tracking2,247
dlr-rm/blenderprocAutomates photorealistic rendering of 3D scenes using Blender2,872
timzhang642/3d-machine-learningA resource repository for 3D machine learning, providing a centralized platform for research papers, datasets, models, and courses related to 3D computer vision and machine learning.9,813
nvidiagameworks/kaolinA PyTorch library for accelerating 3D deep learning research with various GPU-optimized operations and tools.4,550
kubeflow/katibAn automated machine learning system that supports hyperparameter tuning and neural architecture search on Kubernetes.1,521
trekhleb/homemade-machine-learningPractices implementing popular machine learning algorithms from scratch to gain a deeper understanding of their mathematics23,191
sjchoi86/tensorflow-101A comprehensive tutorial project on building and training neural networks with TensorFlow2,606
isl-org/open3dA C++ library for rapid development of software that processes 3D data11,648
kubeflow-kale/kaleSimplifies the deployment of Kubeflow Pipelines workflows by providing a graphical interface for Data Scientists to define and deploy pipelines directly from JupyterLab.632
binroot/tensorflow-bookA comprehensive resource for learning machine learning using TensorFlow.4,453
yunishi3/3d-fcr-alphaganThis project aims to develop a generative model for 3D multi-object scenes using a novel network architecture inspired by auto-encoding and generative adversarial networks.103
google-deepmind/mujocoA general-purpose physics simulator for simulating articulated structures in motion.8,338
mit-spark/kimera-semanticsTools and methods for real-time 3D reconstruction from 2D data in robotics and computer vision651