Orthogonal Knowledge Refreshing for Domain-Incremental Object Detection
For continual learning in object detection, OKR provides a simple yet effective exemplar-free solution to catastrophic forgetting, outperforming prior methods significantly.
OKR addresses inter-domain interference in domain-incremental object detection by constructing independent domain-specific subspaces with low-rank branches and using gradient-based orthogonal refreshing to prevent forgetting, achieving +5.6% and +6.5% mAP gains over exemplar-free methods on Pascal VOC and BDD100K.
Domain-incremental object detection (DIOD) requires models to continually adapt to new domains while preserving prior knowledge. Recently, parameter-efficient fine-tuning offers a promising avenue, wherein a pre-trained model is frozen and a small number of learnable parameters are injected for downstream tasks. However, these methods risk overwriting critical past knowledge, triggering inter-domain interference and performance degradation. To address this challenge, we propose Orthogonal Knowledge Refreshing (OKR), a simple yet effective framework for DIOD. OKR incrementally constructs independent domain-specific subspaces via dedicated low-rank branches for each domain, which are seamlessly fused for a holistic decision, enabling conflict-free capacity expansion without domain selection during inference. To minimize knowledge interference during fusion, we present a gradient-based orthogonal refreshing strategy that projects gradient updates of new domains onto the orthogonal complement of the fused historical subspace, supporting continual adaptation without forgetting. Moreover, to mitigate semantic fragmentation across domains, we enforce topology-aware consistency, aligning the semantic structures of old and new domains. Extensive experiments validate the superiority of OKR, outperforming the best exemplar-free method by significant margins of +5.6% and +6.5% mAP on the Pascal VOC and BDD100K series, respectively.