BlenderProc:CNN学習用のリアルな画像を生成するモジュラーパイプライン
BlenderProc
Blenderを拡張し、セグメンテーションや深度・法線・姿勢推定などのCNN学習に使えるリアルな合成画像を生成するモジュール式パイプラインを提案した。
著者: Maximilian Denninger, Martin Sundermeyer, Dominik Winkelbauer, Youssef Zidan, Dmitry Olefir, Mohamad Elbadrawy, Ahsan Lodhi, Harinandan Katam
分類: cs.CV, cs.GR, cs.LG, cs.RO
原文アブストラクト
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.
