Unsupervised Domain Adaptive Object Detection Using Forward-Backward Cyclic Adaptation

We present a novel approach to perform the unsupervised domain adaptation for object detection through forward-backward cyclic (FBC) training. Recent adversarial training based domain adaptation methods have shown their effectiveness on minimizing domain discrepancy via marginal feature distributions alignment. However, aligning the marginal feature distributions does not guarantee the alignment of class conditional distributions. This limitation is more evident when adapting object detectors as the domain discrepancy is larger compared to the image classification task, e.g., various number of objects exist in one image and the majority of content in an image is the background. This motivates us to learn domain-invariance for category-level semantics via gradient alignment for instance-level adaptation. Intuitively, if the gradients of two domains point in similar directions, then the learning of one domain can improve that of another domain. We propose Forward- Backward Cyclic Adaptation to achieve gradient alignment, which iteratively computes adaptation from source to target via backward hopping and from target to source via forward passing. In addition, we align low-level features for adapting image-level color/texture via adversarial training. However, the detector that performs well on both domains is not ideal for the target domain. As such, in each cycle, domain diversity is enforced by two regularizations: 1) maximum entropy regularization on the source domain to penalize confident source-specific learning and 2) minimum entropy regularization on target domain to intrigue target-specific learning. Theoretical analysis of the training process is provided, and extensive experiments on challenging cross-domain object detection datasets have shown our approach’s superiority over the state-of-the-art.