Abstract:Conventional object detectors often encounter remarkable performance drops due to the domain shift caused by environmental changes. However, labeling sufficient training data drawn from various domains is cost-ineffective and labor-intensive. To this end, unsupervised domain adaptive object detection (DAOD) has attracted much attention, in which the detector is transferred from the label-rich source domain to a label-agnostic target domain. Most of the existing cross-domain object detectors are designed with a parameter-shared network architecture, which, however, has an inherent flaw to accumulate the source errors caused by the inaccurate target distribution. As a result, the target data instead of decisive source data may dominate the learning towards model collapse. To overcome the risk, we propose an Asymmetric Tri-way Faster-RCNN (ATF) for DAOD tasks, in which a novel ancillary net is deployed with the chief net and it brings two merits. 1) The ancillary net enables the source distribution to be preserved without accumulating source error/distortion, such that the target distribution is dominated by source data and the model collapse is alleviated. 2) Ancillary target features can be generated by the ancillary net, which further enhances the discrimination of the target domain object detector. Furthermore, in order to remove the useless source-specific knowledge and exploit the informative domain-invariant knowledge, we propose a Partial Alignment based ATF (PA-ATF) model, in which an adversarial shuffling based mutual-information minimization strategy for domain disentanglement is provided. Intuitively, transferring only the source-domain invariant information to the target domain is more conscionable. Extensive experiments on benchmark datasets, including the Cityscapes, Foggy Cityscapes, Pascal VOC, Clipart, Watercolor, SIM10K, and KITTI demonstrate the remarkable performance of our models over other state-of-the-art approaches.

Cross-domain Object Detection by Local to Global Object-Aware Feature Alignment

Joint Feature-Level And Pixel-Level Domain Adaption For Object Detection In The Wild

Cross-domain Object Detection through Coarse-to-Fine Feature Adaptation

Local–Global Attentive Adaptation for Object Detection

AFAN: Augmented Feature Alignment Network for Cross-Domain Object Detection

Loose to compact feature alignment for domain adaptive object detection

Deeply Aligned Adaptation for Cross-domain Object Detection

Multi-Granularity Alignment Domain Adaptation for Object Detection

H2FA R-CNN: Holistic and Hierarchical Feature Alignment for Cross-domain Weakly Supervised Object Detection

Spatial Alignment for Unsupervised Domain Adaptive Single-Stage Object Detection

Multi-Level Domain Adaptive Learning For Cross-Domain Detection

Robust Domain Adaptive Object Detection with Unified Multi-Granularity Alignment

Domain Adaptive Object Detection via Feature Separation and Alignment

Cross-Domain Object Detection by Dual Adaptive Branch

Cross Domain Object Detection via Multi-Granularity Confidence Alignment based Mean Teacher

NLFA: A Non Local Fusion Alignment Module for Multi-Scale Feature in Object Detection

Align Deep Features for Oriented Object Detection

MLFA: Toward Realistic Test Time Adaptive Object Detection by Multi-Level Feature Alignment

AlignYOLO: A feature-aligned network for object detection

Partial Alignment for Object Detection in the Wild

Adversarial Alignment for Source Free Object Detection