Enhancement of 3D Camera Synthetic Training Data with Noise Models
This work addresses the challenge of domain gap in 3D vision for researchers and practitioners, but it is incremental as it builds on existing noise modeling techniques.
The paper tackled the problem of improving 3D camera synthetic training data by modeling noise from real scanners and applying it to synthetic data, finding that optimal noise levels enhance neural network generalization for object segmentation on real data.
The goal of this paper is to assess the impact of noise in 3D camera-captured data by modeling the noise of the imaging process and applying it on synthetic training data. We compiled a dataset of specifically constructed scenes to obtain a noise model. We specifically model lateral noise, affecting the position of captured points in the image plane, and axial noise, affecting the position along the axis perpendicular to the image plane. The estimated models can be used to emulate noise in synthetic training data. The added benefit of adding artificial noise is evaluated in an experiment with rendered data for object segmentation. We train a series of neural networks with varying levels of noise in the data and measure their ability to generalize on real data. The results show that using too little or too much noise can hurt the networks' performance indicating that obtaining a model of noise from real scanners is beneficial for synthetic data generation.