PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency Training
This addresses the problem of costly annotation in computer vision by enabling models trained on simulated data to perform well on real-world domains, though it is incremental as it builds on consistency training ideas.
The paper tackles unsupervised domain adaptation for semantic segmentation by proposing a pixelwise consistency training framework, achieving strong results on synthetic-to-real benchmarks like GTA5-to-Cityscapes and SYNTHIA-to-Cityscapes.
Unsupervised domain adaptation is a promising technique for semantic segmentation and other computer vision tasks for which large-scale data annotation is costly and time-consuming. In semantic segmentation, it is attractive to train models on annotated images from a simulated (source) domain and deploy them on real (target) domains. In this work, we present a novel framework for unsupervised domain adaptation based on the notion of target-domain consistency training. Intuitively, our work is based on the idea that in order to perform well on the target domain, a model's output should be consistent with respect to small perturbations of inputs in the target domain. Specifically, we introduce a new loss term to enforce pixelwise consistency between the model's predictions on a target image and a perturbed version of the same image. In comparison to popular adversarial adaptation methods, our approach is simpler, easier to implement, and more memory-efficient during training. Experiments and extensive ablation studies demonstrate that our simple approach achieves remarkably strong results on two challenging synthetic-to-real benchmarks, GTA5-to-Cityscapes and SYNTHIA-to-Cityscapes. Code is available at: https://github.com/lukemelas/pixmatch