# Comparative Analysis of Deep Learning Models for Brand Logo Classification in Real-World Scenarios

arXiv: https://arxiv.org/abs/2305.12242 · HTML: https://mlsanity.com/paper/2305.12242/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 11 references that are in the map.
- First posted on arXiv 20 May 2023.

## Builds on (most cited first, 8 of 8)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Densely Connected Convolutional Networks](https://mlsanity.com/paper/1608.06993/) (2016)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [Searching for MobileNetV3](https://mlsanity.com/paper/1905.02244/) (2019)
- [AutoAugment: Learning Augmentation Policies from Data](https://mlsanity.com/paper/1805.09501/) (2018)
- [Image Segmentation Using Deep Learning: A Survey](https://mlsanity.com/paper/2001.05566/) (2020)
- [MultiMAE: Multi-modal Multi-task Masked Autoencoders](https://mlsanity.com/paper/2204.01678/) (2022)
- [DaViT: Dual Attention Vision Transformers](https://mlsanity.com/paper/2204.03645/) (2022)

## Built on by (most cited first, 1 of 1)

- [LU-500: A Logo Benchmark for Concept Unlearning](https://mlsanity.com/paper/2607.24101/) (2026)

## Abstract (arXiv)

This report presents a comprehensive study on deep learning models for brand logo classification in real-world scenarios. The dataset contains 3,717 labeled images of logos from ten prominent brands. Two types of models, Convolutional Neural Networks (CNN) and Vision Transformer (ViT), were evaluated for their performance. The ViT model, DaViT small, achieved the highest accuracy of 99.60%, while the DenseNet29 achieved the fastest inference speed of 366.62 FPS. The findings suggest that the DaViT model is a suitable choice for offline applications due to its superior accuracy. This study demonstrates the practical application of deep learning in brand logo classification tasks.