Scalable Logo Detection by Self Co-Learning.

Hang Su,Shaogang Gong,Xiatian Zhu

DOI: https://doi.org/10.1016/j.patcog.2019.107003

IF: 8

2019-01-01

Pattern Recognition

Abstract:•The literature lacks large-scale logo detection test benchmarks due to rather expensive data selection and label annotation.•We contribute a large-scale dataset collected automatically for scalable logo detection.•We present a scalable logo detection solution characterised by joint co-learning and self-learning in a unified framework, without the tedious need for manually labelling any training data.