Unsupervised RGB-to-Thermal Domain Adaptation via Multi-Domain Attention Network
This work enables robots to perform visual tasks at night and in adverse weather conditions without additional labeling costs, addressing a domain-specific problem with incremental improvements.
The paper tackles unsupervised thermal image classification and segmentation by transferring knowledge from RGB to thermal domains without requiring thermal annotations or co-registered pairs, achieving state-of-the-art performance in classification benchmarks and successfully applying the method to thermal river scene segmentation using only synthetic RGB images.
This work presents a new method for unsupervised thermal image classification and semantic segmentation by transferring knowledge from the RGB domain using a multi-domain attention network. Our method does not require any thermal annotations or co-registered RGB-thermal pairs, enabling robots to perform visual tasks at night and in adverse weather conditions without incurring additional costs of data labeling and registration. Current unsupervised domain adaptation methods look to align global images or features across domains. However, when the domain shift is significantly larger for cross-modal data, not all features can be transferred. We solve this problem by using a shared backbone network that promotes generalization, and domain-specific attention that reduces negative transfer by attending to domain-invariant and easily-transferable features. Our approach outperforms the state-of-the-art RGB-to-thermal adaptation method in classification benchmarks, and is successfully applied to thermal river scene segmentation using only synthetic RGB images. Our code is made publicly available at https://github.com/ganlumomo/thermal-uda-attention.