Search papers, labs, and topics across Lattice.
This paper introduces Socialized Detector Learning (SDL), a novel framework that enhances heterogeneous object detectors through a method called Trajectory-Guided and Reciprocal Distillation (TGRD). By estimating Inter-Detector Transfer Difficulty (IDTD) and constructing a strategic knowledge transfer trajectory, SDL enables effective knowledge sharing among specialized detectors, resulting in improved performance on previously unsupported categories. The approach demonstrates a significant 2.6 AP improvement over traditional aggregation methods on the MS COCO dataset, showcasing the efficacy of order-aware knowledge consolidation and reciprocal transfer in detector evolution.
Knowledge transfer among heterogeneous object detectors can yield up to a 2.6 AP improvement, revolutionizing how we approach multi-teacher distillation in AI.
Object detection knowledge is fragmented across independently trained, heterogeneous detectors with complementary category supports. In socialized learning, this knowledge resides in a society, and learning aims to evolve the society collectively through exchange. However, aggregation-based socialization does not explicitly plan transfer order, whereas progressive multi-teacher distillation considers order but remains a one-way student enhancement in a shared category space. Building on Socialized Learning, we formulate Socialized Detector Learning (SDL) for heterogeneous, category-specialized object detectors and propose Trajectory-Guided and Reciprocal Distillation (TGRD).TGRD estimates directed operational Inter-Detector Transfer Difficulty (IDTD) from held-out feature-alignment residuals, precomputes a fixed score table, and greedily constructs a carrier trajectory. Along the trajectory, knowledge is progressively consolidated into a union-category carrier and then returned to experts through reciprocal transfer. A conditional proxy-certificate analysis shows that, under stated assumptions, the progressive certificate is no larger than an aggregated-target counterpart. On MS COCO with four heterogeneous experts and two carrier initializations, final carriers outperform epoch-matched simultaneous aggregation controls by 2.6 AP in both settings. Reciprocal detectors attain 20.8--28.4 AP on previously unsupported categories while remaining within 1.3 AP of original expert-specific performance. These results support order-aware progressive consolidation followed by reciprocal transfer as a viable mechanism for detector-society evolution.
