BlenderProc

2019/10/25 by Maximilian Denninger, Denninger, Maximilian, Martin Sundermeyer +13 · 7 citations
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Graphics (cs.GR) #Machine Learning (cs.LG) #Robotics (cs.RO) #cs.CV #cs.GR #cs.LG #cs.RO

paper · pdf · doi:10.48550/arxiv.1911.01911

7 pages, 8 figures

arxiv created 2019/10/25 · arxiv updated 2019/11/06

Abstract

BlenderProc is a modular procedural pipeline, which helps in generating real looking images for the training of convolutional neural networks. These can be used in a variety of use cases including segmentation, depth, normal and pose estimation and many others. A key feature of our extension of blender is the simple to use modular pipeline, which was designed to be easily extendable. By offering standard modules, which cover a variety of scenarios, we provide a starting point on which new modules can be created.

Cited by

• Object Pose Distribution Estimation for Determining Revolution and Reflection Uncertainty in Point Clouds
• WALDO: Where Unseen Model-based 6D Pose Estimation Meets Occlusion
• SCOPE: Semantic Conditioning for Sim2Real Category-Level Object Pose Estimation in Robotics
• A Synthetic Dataset for Manometry Recognition in Robotic Applications
• Geometry-Aware Global Feature Aggregation for Real-Time Indirect Illumination
• AI-driven visual monitoring of industrial assembly tasks
• Sim2Real Transfer for Vision-Based Grasp Verification

Related

• EasyPBR: A Lightweight Physically-Based Renderer
• Soft Rasterizer: Differentiable Rendering for Unsupervised Single-View Mesh Reconstruction
• BlendTorch: A Real-Time, Adaptive Domain Randomization Library
• SMPLpix: Neural Avatars from 3D Human Models
• PyTorchPipe: a framework for rapid prototyping of pipelines combining\n language and vision