ECAD: Expanding Class-Agnostic Detection Beyond Thing-Centric Objectness

2026-08-07

AI Summary

Traditional object detection is constrained by the “object-centric” paradigm, which assumes well-defined boundaries and struggles to capture unbounded yet semantically critical visual elements such as sky or road. To address this limitation, this work proposes Extended Category-Agnostic Detection (ECAD), a novel paradigm for detecting generalized visual elements. We introduce BTCO-Bench, the first benchmark tailored for such elements, and present ECADet, a lightweight detector built upon a frozen DINOv3 encoder and DETR architecture. ECADet incorporates Geometry-Aware Expert Regression (GAER) and Prototype-Guided Query Modulation (PGQM) to enhance objectness estimation and localization across diverse visual content. Experiments demonstrate that ECADet significantly outperforms existing methods on BTCO-Bench, validating the efficacy of extended objectness modeling.

Abstract

Object detection is a fundamental task in visual perception, providing structured region representations for recognition, grounding, reasoning, and interaction. However, existing detection paradigms largely inherit a thing-centric notion of objectness, where detectors are mainly trained to localize discrete and countable object instances. Consequently, many semantically meaningful visual elements, such as sky, road, grassland, water, and sports courts, are often absorbed into the background despite their importance for scene understanding and spatial reasoning. In this paper, we formulate Expanded Class-Agnostic Detection (ECAD), a new setting that aims to discover category-agnostic visual candidates beyond conventional thing-centric objects. To support this setting, we construct BTCO-Bench, a Beyond Thing-Centric Objectness benchmark with category-agnostic box annotations covering both real-world and cross-domain scenarios. We further propose ECADet, a lightweight DETR-based detector built upon a frozen DINOv3 encoder, and introduce Geometry-Aware Expert Regression (GAER) and Prototype-Guided Query Modulation (PGQM) to improve localization and objectness estimation for diverse visual elements, respectively. Extensive experiments show that ECADet consistently outperforms representative class-agnostic and proposal-based detectors on BTCO-Bench, demonstrating the effectiveness of expanded objectness discovery. Code and benchmark will be released.

Problem

• class-agnostic detection
• thing-centric objectness
• visual elements
• scene understanding
• objectness discovery

Innovation

• Expanded Class-Agnostic Detection
• Beyond Thing-Centric Objectness
• Geometry-Aware Expert Regression
• Prototype-Guided Query Modulation
• BTCO-Bench