Order Matters: Competition-Guided Query Ordering for RNN-Based Object Detection

2026-10-04

AI Summary

This study addresses the limitations of DETR arising from the absence of an explicit deduplication mechanism, which leads to query redundancy, training instability, and prediction uncertainty. To this end, it proposes DETRNN, a plug-and-play module inspired by Non-Maximum Suppression (NMS) that introduces confidence- and similarity-based explicit ranking. This approach transforms unordered object queries into competition-aware sequences, enabling recursive refinement via a recurrent neural network along the ordered sequence to explicitly model and optimize inter-query competitive relationships within the decoder. The proposed module can be seamlessly integrated into various DETR architectures, yielding consistent accuracy improvements while maintaining computational efficiency. Ultimately, DETRNN effectively suppresses redundant predictions and stabilizes the optimization process.

Abstract

DETR-style detectors use one-to-one bipartite matching during training to assign object queries to ground-truth objects, enabling end-to-end set prediction without non-maximum suppression (NMS). However, without an explicit de-duplication procedure, multiple queries can still produce highly similar hypotheses for the same object, making training unstable and predictions less decisive. Inspired by the sequential ordering of NMS, we propose DETRNN, a plug-and-play module that turns unordered object queries into a competition-aware sequence for recurrent refinement. DETRNN builds an explicit confidence-and-similarity based order from prior predictions, then refines queries with an RNN along this order to model competition inside the decoder. This ordered recurrent refinement reduces redundant predictions, stabilizes optimization, and improves final detection accuracy. Experiments on multiple DETR-style detectors show consistent gains with comparable efficiency.

Problem

• Object Detection
• DETR
• Query Redundancy
• Training Instability
• Bipartite Matching

Innovation

• DETR
• Object Detection
• Query Ordering
• RNN
• Competition-Guided Refinement