Spatial Attention Pyramid Network for Unsupervised Domain Adaptation

By Congcong Li, Dawei Du, Libo Zhang, Longyin Wen, Tiejian Luo, Yanjun Wu, Pengfei Zhu. 2020

135 citations recorded by the source corpus.

Unsupervised domain adaptation is critical in various computer vision tasks, such as object detection, instance segmentation, and semantic segmentation, which aims to alleviate performance degradation caused by domain-shift. Most of previous methods rely on a single-mode distribution of source and target domains to align them with adversarial learning, leading to inferior results in various scenarios. To that end, in this paper, we design a new spatial attention pyramid network for unsupervised domain adaptation. Specifically, we first build the spatial pyramid representation to capture context information of objects at different scales. Guided by the task-specific information, we combine the dense global structure representation and local texture patterns at each spatial location effectively using the spatial attention mechanism. In this way, the network is enforced to focus on the discriminative regions with context information for domain adaption. We conduct extensive experiments on various challenging datasets for unsupervised domain adaptation on object detection, instance segmentation, and semantic segmentation, which demonstrates that our method performs favorably against the state-of-the-art methods by a large margin. Our source code is available at this https URL.

Subfields: Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

Citations in the corpus

Towards Real-World Prohibited Item Detection: A Large-Scale X-ray Benchmark (2021, background): “In recent years, due to the substantial development of deep learning and computer vision technologies[31, 24, 34, 14, 13, 16, 3], automatic security inspection of prohibited items becomes possible.”
MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation (2022, methodology): “To facilitate learning domain-robust context dependencies, several UDA methods propose network components that can capture context information such as spatial attention pyramids [47], cross-domain attention [103], or context-aware feature fusion [34].”
Visual Attention Methods in Deep Learning: An In-Depth Survey (2022, methodology): “[39] introduced a spatial attention pyramid that takes features from multiple average pooling layers”
Unsupervised Domain Adaptation in Semantic Segmentation via Orthogonal and Clustered Embeddings (2020, methodology): “When compared with other approaches, our method performs better than standard feature-level adversarial techniques [22, 21, 57, 25].”