Domain Adaptation for Camera-Specific Image Characteristics using Shallow Discriminators

TL;DR

This paper introduces shallow discriminator architectures for camera-specific image domain adaptation, improving local distortion modeling and achieving higher segmentation accuracy with significantly reduced model complexity.

Contribution

It proposes shallow discriminator designs that enhance local distortion learning and demonstrate superior efficiency and accuracy in camera-specific image domain adaptation tasks.

Findings

Up to 0.15 mean average precision improvement for distortions

Up to 0.16 improvement for camera-specific characteristics

Reduces model complexity by a factor of 20 compared to some methods

Abstract

Each image acquisition setup leads to its own camera-specific image characteristics degrading the image quality. In learning-based perception algorithms, characteristics occurring during the application phase, but absent in the training data, lead to a domain gap impeding the performance. Previously, pixel-level domain adaptation through unpaired learning of the pristine-to-distorted mapping function has been proposed. In this work, we propose shallow discriminator architectures to address limitations of these approaches. We show that a smaller receptive field size improves learning of unknown image distortions by more accurately reproducing local distortion characteristics at a low network complexity. In a domain adaptation setup for instance segmentation, we achieve mean average precision increases over previous methods of up to 0.15 for individual distortions and up to 0.16 for camera-specific image characteristics in a simplified camera model. In terms of number of parameters, our approach matches the complexity of one state of the art method while reducing complexity by a factor of 20 compared to another, demonstrating superior efficiency without compromising performance.