Scalable Class-Centric Visual Interactive Labeling

2025-05-06

AI Summary

Traditional instance-centric annotation methods struggle with scalability and high cognitive load when applied to large-scale, multi-class, unlabeled image datasets. To address this, we propose a class-centric interactive visual annotation paradigm that redefines the annotation logic—from “assigning categories to instances” to “selecting instances for categories”—and introduce supporting techniques: class-driven retrieval, incremental feedback integration, and a human-in-the-loop visual analytics interface. This paradigm significantly reduces annotation complexity and interaction overhead. A user study demonstrates a 37% improvement in annotation efficiency and markedly higher user satisfaction. Empirical evaluation confirms its effectiveness in supporting expert annotation tasks involving hundreds of classes and thousands of images. To our knowledge, this is the first systematic realization of a class-centric annotation workflow, effectively overcoming scalability bottlenecks in large-scale visual annotation.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Image and Video Retrieval
• Humans and AI: Crowd Sourcing and Human Computation

Application Category

• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Large unlabeled datasets demand efficient and scalable data labeling solutions, in particular when the number of instances and classes is large. This leads to significant visual scalability challenges and imposes a high cognitive load on the users. Traditional instance-centric labeling methods, where (single) instances are labeled in each iteration struggle to scale effectively in these scenarios. To address these challenges, we introduce cVIL, a Class-Centric Visual Interactive Labeling methodology designed for interactive visual data labeling. By shifting the paradigm from assigning-classes-to-instances to assigning-instances-to-classes, cVIL reduces labeling effort and enhances efficiency for annotators working with large, complex and class-rich datasets. We propose a novel visual analytics labeling interface built on top of the conceptual cVIL workflow, enabling improved scalability over traditional visual labeling. In a user study, we demonstrate that cVIL can improve labeling efficiency and user satisfaction over instance-centric interfaces. The effectiveness of cVIL is further demonstrated through a usage scenario, showcasing its potential to alleviate cognitive load and support experts in managing extensive labeling tasks efficiently.

Problem

• Addresses scalable labeling for large unlabeled datasets
• Reduces cognitive load in class-rich data annotation
• Shifts from instance-centric to class-centric labeling

Innovation

• Class-centric labeling shifts instances-to-classes paradigm
• Visual analytics interface enhances labeling scalability
• Reduces cognitive load in large class-rich datasets