Knowing the Unknown: Interpretable Open-World Object Detection via Concept Decomposition Model

2026-02-24

AI Summary

This work addresses the challenges of unreliable unknown object recognition and poor interpretability in open-world object detection, where known and unknown categories are often confused. To this end, we propose IPOW, an interpretable open-world object detection framework that introduces concept-level interpretability for the first time. IPOW employs a Concept Decomposition Model (CDM) to disentangle RoI features into discriminative, shared, and background concepts, and incorporates a Concept-Guided Rectification (CGR) mechanism to mitigate category confusion. Built upon the Faster R-CNN architecture, IPOW significantly improves recall for unknown classes, effectively reduces confusion between known and unknown categories, and provides concept-level interpretability to support detection decisions.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Calibration & Uncertainty Quantification
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Open-world object detection (OWOD) requires incrementally detecting known categories while reliably identifying unknown objects. Existing methods primarily focus on improving unknown recall, yet overlook interpretability, often leading to known-unknown confusion and reduced prediction reliability. This paper aims to make the entire OWOD framework interpretable, enabling the detector to truly"knowing the unknown". To this end, we propose a concept-driven InterPretable OWOD framework(IPOW) by introducing a Concept Decomposition Model (CDM) for OWOD, which explicitly decomposes the coupled RoI features in Faster R-CNN into discriminative, shared, and background concepts. Discriminative concepts identify the most discriminative features to enlarge the distances between known categories, while shared and background concepts, due to their strong generalization ability, can be readily transferred to detect unknown categories. Leveraging the interpretable framework, we identify that known-unknown confusion arises when unknown objects fall into the discriminative space of known classes. To address this, we propose Concept-Guided Rectification (CGR) to further resolve such confusion. Extensive experiments show that IPOW significantly improves unknown recall while mitigating confusion, and provides concept-level interpretability for both known and unknown predictions.

Problem

• open-world object detection
• unknown object identification
• interpretability
• known-unknown confusion

Innovation

• Interpretable OWOD
• Concept Decomposition Model
• Concept-Guided Rectification
• Open-World Object Detection
• Feature Disentanglement