Contrastive Multi-View Textual-Visual Encoding: Towards One Hundred Thousand-Scale One-Shot Logo Identification

TL;DR

This paper introduces a contrastive multi-view encoding framework for large-scale one-shot logo identification, leveraging textual and visual features, and presents a new dataset with 100K logos to advance research in open-set recognition.

Contribution

The paper proposes a novel multi-view textual-visual encoding method for one-shot logo recognition and introduces WiRLD, a large-scale logo dataset with 100K logos, addressing scalability and dataset gaps.

Findings

Achieves 91.3% AUC on QMUL-OpenLogo verification

Outperforms state-of-the-art by 9.1% and 2.6% on Toplogos-10 and FlickrLogos32

More stable at 100K candidate logos scale

Abstract

In this paper, we study the problem of identifying logos of business brands in natural scenes in an open-set one-shot setting. This problem setup is significantly more challenging than traditionally-studied 'closed-set' and 'large-scale training samples per category' logo recognition settings. We propose a novel multi-view textual-visual encoding framework that encodes text appearing in the logos as well as the graphical design of the logos to learn robust contrastive representations. These representations are jointly learned for multiple views of logos over a batch and thereby they generalize well to unseen logos. We evaluate our proposed framework for cropped logo verification, cropped logo identification, and end-to-end logo identification in natural scene tasks; and compare it against state-of-the-art methods. Further, the literature lacks a 'very-large-scale' collection of reference logo images that can facilitate the study of one-hundred thousand-scale logo identification. To fill this gap in the literature, we introduce Wikidata Reference Logo Dataset (WiRLD), containing logos for 100K business brands harvested from Wikidata. Our proposed framework that achieves an area under the ROC curve of 91.3% on the QMUL-OpenLogo dataset for the verification task, outperforms state-of-the-art methods by 9.1% and 2.6% on the one-shot logo identification task on the Toplogos-10 and the FlickrLogos32 datasets, respectively. Further, we show that our method is more stable compared to other baselines even when the number of candidate logos is on a 100K scale.