DeepLogo: Hitting Logo Recognition with the Deep Neural Network Hammer

TL;DR

This paper applies deep convolutional neural networks to logo recognition, proposing new architectures that outperform previous methods on a standard dataset, advancing the state-of-the-art in this domain.

Contribution

Introduction of novel DCNN architectures specifically designed for logo recognition, achieving superior accuracy over existing approaches.

Findings

Surpassed state-of-the-art accuracy on a popular logo dataset

Demonstrated effectiveness of DCNNs in logo recognition tasks

Provided insights into architecture design for logo recognition

Abstract

Recently, there has been a flurry of industrial activity around logo recognition, such as Ditto's service for marketers to track their brands in user-generated images, and LogoGrab's mobile app platform for logo recognition. However, relatively little academic or open-source logo recognition progress has been made in the last four years. Meanwhile, deep convolutional neural networks (DCNNs) have revolutionized a broad range of object recognition applications. In this work, we apply DCNNs to logo recognition. We propose several DCNN architectures, with which we surpass published state-of-art accuracy on a popular logo recognition dataset.