Adapting Object Detectors via Selective Cross-Domain Alignment

By Xinge Zhu, Jiangmiao Pang, Ceyuan Yang, Jianping Shi, Dahua Lin. Computer Vision and Pattern Recognition, 2019

378 citations recorded by the source corpus.

State-of-the-art object detectors are usually trained on public datasets. They often face substantial difficulties when applied to a different domain, where the imaging condition differs significantly and the corresponding annotated data are unavailable (or expensive to acquire). A natural remedy is to adapt the model by aligning the image representations on both domains. This can be achieved, for example, by adversarial learning, and has been shown to be effective in tasks like image classification. However, we found that in object detection, the improvement obtained in this way is quite limited. An important reason is that conventional domain adaptation methods strive to align images as a whole, while object detection, by nature, focuses on local regions that may contain objects of interest. Motivated by this, we propose a novel approach to domain adaption for object detection to handle the issues in ``where to look'' and ``how to align''. Our key idea is to mine the discriminative regions, namely those that are directly pertinent to object detection, and focus on aligning them across both domains. Experiments show that the proposed method performs remarkably better than existing methods with about 4% ~ 6% improvement under various domain-shift scenarios while keeping good scalability.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/a9aee599c8ca734cfdf071830fcd636f60a92fbf

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Unbiased Mean Teacher for Cross-domain Object Detection (2020, methodology): “Besides the baselines compared previously, we further include DA-Faster [6], PF [40], SCDA [56], MAF [17], WD [52], MTOR [1], iFAN [57] and CDN [44] for comparison.”
Progressive Domain Adaptation for Object Detection (2019, methodology): “Table 3 shows that our method reduces the domain gap across weather conditions and performs favorably against the state-of-the-art methods [3, 25, 37, 16].”
Exploring Categorical Regularization for Domain Adaptive Object Detection (2020, methodology): “Thanks to the flexibility of Faster R-CNN, recently, it is widely adapted for domain adaptive object detection [1, 28, 41, 12] with adversarial training [5].”
Cross-Domain Adaptive Teacher for Object Detection (2021, methodology): “Researchers have proposed various methods, such as domain classifier and adversarial learning [10], to address the task of cross-domain adaptation in object detection [2, 3, 14, 32, 39, 42, 44].”
Spatial Attention Pyramid Network for Unsupervised Domain Adaptation (2020, methodology): “6 SCDA [45] 33.”
A Review of Single-Source Deep Unsupervised Visual Domain Adaptation (2020, background): “homogeneity homogeneous DA dS = dT source and target data are observed in the same space [11, 23, 36, 44–106, 113–151] heterogeneous DA dS 6= dT source and target data are observed in different spaces [107–112]”
Libra R-CNN: Towards Balanced Learning for Object Detection (2019, methodology): “Other methods focus on cascade procedures [24], duplicate removal [14, 13], multi-scales [2, 1, 31, 30], adversarial learning [37] and more contextual [36].”
MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation (2022, methodology): “Due to the ubiquity of domain gaps, UDA methods were designed for all major computer vision problems including image classification [20, 21, 57–59, 66, 68, 74, 82, 108], semantic segmentation [1,25,31,32,34,35,85,86,89,92,106], and object detection [4, 7, 8, 15, 50, 53, 73, 94, 100, 113].”
Adapting Object Detectors with Conditional Domain Normalization (2020, methodology): “d 3D ROIs proposal stage (instance alignment). 5 Experimental Results and Analysis 5.1 Results on Cityscapes to Foggy Cityscapes We compare CDN with the state-of-the-art methods in Table 1. Following [34, 47], we also report results using Faster R-CNN model with VGG16 backbone. As shown in Table 1, ”
Harmonizing Transferability and Discriminability for Adapting Object Detectors (2020, methodology): “Instance-level alignment refers to the ROI-Pooling based feature alignment, which has been explored by some prior efforts [7, 62, 18].”