Simple Supervision Is Hard to Beat: A Bitter Lesson from Sparse Target Labels in Domain-Adaptive Object Detection
For practitioners of domain-adaptive object detection, this work demonstrates that simple supervised fine-tuning with sparse labels is more effective than sophisticated pseudo-labeling strategies.
The paper shows that adding a small number (1-10%) of randomly sampled labeled target images to source-free domain adaptive object detection, via a simple supervised loss (RTSM), consistently improves performance by 1.7-18.3 AP50 over pure self-training, while more complex feedback plugins yield limited gains.
Source-free domain adaptive object detection adapts a source-trained detector to an unlabeled target domain, typically through teacher-student self-training with pseudo-labels. We revisit this setting when a small, uniformly sampled subset of target images is labeled. We introduce Random-Target Supervised Mixing (RTSM), a simple anchor that incorporates these annotations through a supervised detection loss while leaving the original unlabeled adaptation branch unchanged. Across evaluations spanning four SFDA-OD methods, two object detectors, multiple adaptation tasks, and target-label budgets from 1% to 10%, RTSM consistently improves pure SFDA by 1.7 to 18.3 AP50. We then examine whether the same annotations can provide further gains by steering unlabeled self-training. To this end, we evaluate ten sparse-label feedback plugins covering pseudo-label selection, object completion, and optimization control, which yield limited and method-dependent gains over RTSM. These results reveal a bitter lesson for sparse-label SFDA-OD: simple supervision is hard to beat. RTSM therefore provides a simple yet effective anchor for sparse-label SFDA-OD.