CAD2Real: Deep learning with domain randomization of CAD data for 3D pose estimation of electronic control unit housings

TL;DR

This paper presents CAD2Real, a deep learning method that uses domain randomization of synthetic CAD data to accurately estimate the 3D pose of electronic control unit housings in real-world images, overcoming data scarcity issues.

Contribution

It introduces a novel domain randomization technique for training neural networks solely on synthetic CAD data to perform 3D pose estimation of ECU housings in real images.

Findings

Effective pose estimation on real images using synthetic data.

Minimal measurement setup requirements.

Transferability to related use cases.

Abstract

Electronic control units (ECUs) are essential for many automobile components, e.g. engine, anti-lock braking system (ABS), steering and airbags. For some products, the 3D pose of each single ECU needs to be determined during series production. Deep learning approaches can not easily be applied to this problem, because labeled training data is not available in sufficient numbers. Thus, we train state-of-the-art artificial neural networks (ANNs) on purely synthetic training data, which is automatically created from a single CAD file. By randomizing parameters during rendering of training images, we enable inference on RGB images of a real sample part. In contrast to classic image processing approaches, this data-driven approach poses only few requirements regarding the measurement setup and transfers to related use cases with little development effort.