Search papers, labs, and topics across Lattice.
This paper introduces class-geometry supervision (CGS), a framework that enhances prototype-based object detectors by constraining learned class representations to reflect visual and semantic dissimilarities among classes. By aligning pairwise distances in the class representation space with a target class-geometry matrix, CGS improves sample efficiency in various detection tasks, including few-shot biomedical object detection and open-set detection. The experiments demonstrate that CGS significantly boosts performance in novel-class insertion and unknown recall on COCO, while maintaining strong known-class detection capabilities, highlighting the importance of relational class geometry in open-world detection scenarios.
Class-geometry supervision transforms how we approach sample-efficient open-world detection, yielding significant gains in novel-class insertion and unknown recall.
Open-world object detection requires models to recognize known categories, reject unfamiliar objects, and incorporate new classes over time. This is especially challenging in scarce-data settings such as biomedical and scientific imaging, where rare categories may have only a few annotated examples and fine-grained classes differ by subtle morphology. Prototype-based detectors are natural for this regime, but they typically learn class prototypes as independent anchors, ignoring relational structure among classes. We propose class-geometry supervision (CGS), a general framework that constrains learned prototype or class-representation spaces to preserve visual or semantic class dissimilarities estimated from training data. CGS introduces a dissimilarity-preserving objective that aligns pairwise distances among learned class representations with a target class-geometry matrix while retaining the standard task loss. We instantiate the same objective across prototype recognition, few-shot biomedical object detection, open-set detection, novel-class insertion, and OWOD adaptation on COCO. Experiments show that CGS improves sample efficiency in recognition and ova detection, substantially strengthens novel-class insertion, and improves unknown recall on COCO while retaining much of the known-class detection performance. Ablations show that meaningful visual geometry provides the most reliable gains, while random geometry can help novel separation but is less consistent for few-shot detection. These results suggest that relational class geometry is an effective supervisory signal for building calibrated and extensible open-world detectors under limited supervision.
