Domain Generalization with MixStyle

Abstract

Though convolutional neural networks (CNNs) have demonstrated remarkable ability in learning discriminative features, they often generalize poorly to unseen domains. Domain generalization aims to address this problem by learning from a set of source domains a model that is generalizable to any unseen domain. In this paper, a novel approach is proposed based on probabilistically mixing instance-level feature statistics of training samples across source domains. Our method, termed MixStyle, is motivated by the observation that visual domain is closely related to image style (e.g., photo vs.~sketch images). Such style information is captured by the bottom layers of a CNN where our proposed style-mixing takes place. Mixing styles of training instances results in novel domains being synthesized implicitly, which increase the domain diversity of the source domains, and hence the generalizability of the trained model. MixStyle fits into mini-batch training perfectly and is extremely easy to implement. The effectiveness of MixStyle is demonstrated on a wide range of tasks including category classification, instance retrieval and reinforcement learning.

ICLR 2021; Code is available at https://github.com/KaiyangZhou/mixstyle-release

References in corpus (8)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Improved Regularization of Convolutional Neural Networks with Cutout
• A Learned Representation For Artistic Style
• Domain Generalization via Model-Agnostic Learning of Semantic Features
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking
• Torchreid: A Library for Deep Learning Person Re-Identification in Pytorch
• Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck