Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces Symbiosis-Inspired Knowledge Distillation (SIKD) to enhance incremental object detection by addressing the challenges of catastrophic forgetting and class confusion. By leveraging spatial and semantic dependencies through Spatial Symbiosis Distillation (SpSD) and Semantic Symbiosis Distillation (SeSD), the method preserves essential features from old classes while adapting to new categories. Experimental results show that SIKD significantly outperforms existing class incremental learning approaches, leading to improved detection accuracy and knowledge retention.
Ignoring object symbiosis in incremental learning can lead to catastrophic forgetting, but SIKD effectively preserves old knowledge while integrating new classes.
Incremental object detection (IOD) aims to extend detectors to new categories while retaining previously acquired knowledge. Existing methods often adopt a class incremental learning perspective, separating feature spaces to sharpen decision boundaries. However, this separation-oriented paradigm may overlook object symbiosis in detection, where co-occurrence and occlusion introduce spatial and semantic dependencies that benefit from shared representations. Ignoring these dependencies distorts the shared representations, exacerbates confusion between old and new classes, and accelerates catastrophic forgetting. To address this, we propose Symbiosis-Inspired Knowledge Distillation (SIKD), which explicitly leverages object symbiosis at two complementary levels. Spatial Symbiosis Distillation (SpSD) focuses on symbiotic regions where the old model responds with high overlap to objects in the new task. It preserves generalizable old class cues, suppresses class-specific bias and redundancy, and distills the refined evidence to the new model at matched spatial locations with slot-aligned supervision. Semantic Symbiosis Distillation (SeSD) maintains class level structure by forming confidence weighted prototypes for old classes and aligning their inter class soft ranks over the old class logits, which stabilizes the semantic topology during adaptation. Extensive experiments demonstrate the effectiveness and superiority of the proposed method.