FunnelAL: Retrieve-then-Rank Active Learning for Single-Class Discovery
Quick Answer
FunnelAL is a novel active learning system that enhances single-class discovery by using a multi-stage funnel architecture, achieving superior F1 scores and annotation efficiency across three benchmarks.
Quick Take
It outperforms traditional methods like GAL and PF-MA, especially under realistic annotator error rates, by effectively narrowing down candidate samples and refining selection through iterative feedback.
Key Points
- FunnelAL achieves best final F1 scores on three diverse image classification benchmarks.
- It demonstrates the highest annotation efficiency, ranking first in AULC.
- The system uses a three-stage process: retrieval scoring, ranking, and feedback refinement.
- FunnelAL outperforms GAL and PF-MA while reducing annotation rounds.
- Under realistic annotator errors, it maintains top performance compared to classical methods.
Paper Resources
📖 Reader Mode
~2 min readAbstract:We present FunnelAL, a retrieve-then-rank active learning system for single-class discovery, which adapts the multi-stage funnel architecture of industrial recommender systems to data annotation. Large-scale supervised learning faces two challenges: efficiently finding relevant samples in a massive corpus, and distinguishing true positives from visually confusable negatives when embeddings do not cleanly separate classes. Conventional active learning offers a principled framework for reducing annotation cost, yet it treats sample selection as a single-stage process that addresses neither challenge efficiently. FunnelAL decomposes the problem into cascaded stages. Starting from a single positive and negative example, the system iterates through: (1) embedding-based retrieval scoring that narrows the corpus to a manageable candidate set; (2) a precision-triggered ranking stage that exploits a learned ranker (RankNet) while batch precision remains high, then automatically blends in committee-based exploration (QBC) once returns diminish; and (3) feedback from the annotator's labels that refines both stages in subsequent iterations. We evaluate on three diverse image classification benchmarks. With a perfect annotator, FunnelAL attains the best final F1 on all three benchmarks, the best annotation efficiency (first in AULC), and the fewest annotation rounds. The most recent single-class discovery methods (GAL, PF-MA) at best match its final quality, and only at consistently higher labeling cost. Under annotator labeling errors at realistic rates, FunnelAL remains first or statistically tied for first while classical uncertainty-based methods degrade two to three times faster. Our work provides a concrete bridge between multi-stage recommender systems and active learning.
| Comments: | 15 pages, 6 figures, 3 tables |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR); Machine Learning (cs.LG) |
| Cite as: | arXiv:2607.25276 [cs.CV] |
| (or arXiv:2607.25276v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2607.25276 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Brian Goodwin [view email]
[v1]
Tue, 28 Jul 2026 04:26:26 UTC (2,754 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.