Learning Regional Attention over Multi-resolution Deep Convolutional Features for Trademark Retrieval

TL;DR

This paper enhances deep feature aggregation for trademark retrieval by introducing modifications to R-MAC, including multi-resolution inputs, soft-attention, and combined pooling, leading to improved accuracy on large-scale datasets.

Contribution

The authors propose three effective modifications to R-MAC for better handling background clutter, scale variance, and spatial information in trademark retrieval.

Findings

All modifications improve retrieval performance.

The combined approach surpasses previous state-of-the-art results.

Enhancements are validated on the METU dataset.

Abstract

Large-scale trademark retrieval is an important content-based image retrieval task. A recent study shows that off-the-shelf deep features aggregated with Regional-Maximum Activation of Convolutions (R-MAC) achieve state-of-the-art results. However, R-MAC suffers in the presence of background clutter/trivial regions and scale variance, and discards important spatial information. We introduce three simple but effective modifications to R-MAC to overcome these drawbacks. First, we propose the use of both sum and max pooling to minimise the loss of spatial information. We also employ domain-specific unsupervised soft-attention to eliminate background clutter and unimportant regions. Finally, we add multi-resolution inputs to enhance the scale-invariance of R-MAC. We evaluate these three modifications on the million-scale METU dataset. Our results show that all modifications bring non-trivial improvements, and surpass previous state-of-the-art results.