This method scales up existing logo detection models that rely on conventional supervised learning due to no need for large labelled training data per class.
Specifically,. MPCC takes as input the genuine training images of labelled classes, synthetic training images of 1-shot icon supervised classes, and auxiliary ...
This article introduces LogoDet-3K, the largest logo detection dataset with full annotation, which has 3,000 logo categories, about 200,000 manually ...
Multi-perspective cross-class domain adaptation for open logo detection. H Su, S Gong, X Zhu. Computer vision and image understanding 204, 103156, 2021. 14 ...
Dec 18, 2020 · Existing logo detection methods mostly rely on supervised learning with a large quantity of labelled training data in limited classes.
Patern Recognition, Vol. 114, 107862, February 2021 (PR) [ PDF ]; Multi-Perspective Cross-Class Domain Adaptation for Open Logo Detection Hang Su, Shaogang ...
To facilitate efficient cross-domain knowledge transfer without source data, we initialize MFE using parameters of a pretrained source model.
Missing: open | Show results with:open
44 Citations · A Scalable Data Augmentation and Training Pipeline for Logo Detection · Multi-perspective cross-class domain adaptation for open logo detection.
We present Multi-Representation Adaptation Network (MRAN) to accomplish the cross-domain image classification task via multi-representation alignment.
Missing: open | Show results with:open
Jun 3, 2024 · This chapter discusses a domain adaptation-based technique aimed at mitigating the domain-shift challenge in logo detection.