DualMem: Bypassing the Objectness Bottleneck for Calibrated Unknown-Stream Filtering in Open-World Object Detection

2026-05-22

AI Summary

This work addresses the challenge of background-induced false positives in open-world object detection, which hinders the efficiency of incremental learning by corrupting unknown-object predictions. The authors propose DualMem, a non-parametric post-hoc calibration method operating in a frozen SigLIP feature space. By constructing positive and negative memory banks and applying k-nearest neighbors combined with a likelihood ratio test under the Neyman–Pearson framework, DualMem effectively discriminates genuine unknown objects from background noise. The approach uncovers an information bottleneck in the objectness head and introduces, for the first time, a non-parametric calibration mechanism grounded in pretrained visual features. Evaluated on M-OWODB Task 1, DualMem reduces background-type false positives by 56.6% on average without compromising known-class detection performance (mAP remains unchanged), outperforming the K-means baseline by more than twofold.

Abstract

Open-world object detection (OWOD) requires detectors to localize known classes while identifying unknown objects for future incremental learning. We find that the unknown prediction streams of strong OWOD detectors are heavily polluted: on M-OWODB, across PROB, OW-DETR, and HypOW, future-task positive unknowns make up less than 10% of unknown predictions, whereas background false positives account for 46-71%. We show that this is not a missing-information problem, but an information bottleneck at the objectness head. On PROB Task 1, a linear probe on the 256-D decoder query achieves an AUROC of 0.908 for positive-versus-negative unknown discrimination, but the final one-dimensional objectness scalar drops to 0.642. A frozen SigLIP feature, without access to the detector, independently recovers much of this proposal-level separability at the filtering stage (AUROC = 0.871). Motivated by this finding, we propose DualMem, a calibrated post-hoc filter that assumes a small image-disjoint annotated calibration split of held-out future-task objects and performs a non-parametric likelihood ratio test in frozen SigLIP feature space. DualMem uses a k-nearest-neighbor positive memory to protect future-task objects and a negative memory to suppress background-like proposals. Its decision threshold is chosen by Neyman-Pearson calibration, giving users an explicit trade-off between false-unknown suppression and novel recall. Across PROB, OW-DETR, and HypOW on M-OWODB Task 1, DualMem reduces background-type false unknown proposals per image by 44.9%-66.3%, with a mean reduction of 56.6%. On PROB Task 1, it more than doubles the reduction achieved by a natural K-means prototype baseline, while leaving known-class mAP unchanged because known detections bypass the filter.

Problem

• open-world object detection
• unknown-stream filtering
• objectness bottleneck
• false positives
• calibrated filtering

Innovation

• DualMem
• open-world object detection
• objectness bottleneck
• SigLIP features
• non-parametric filtering