MSDU-net: A Multi-Scale Dilated U-net for Blur Detection

TL;DR

This paper introduces MSDU-net, a multi-scale dilated U-net architecture that effectively detects blurred regions in images by leveraging multi-scale texture features, outperforming existing methods on benchmark datasets.

Contribution

The paper proposes a novel multi-scale dilated U-net architecture specifically designed for blur detection, improving segmentation accuracy over previous approaches.

Findings

Outperforms state-of-the-art blur detection methods on benchmark datasets

Utilizes multi-scale dilated convolutions for enhanced texture feature extraction

Achieves better segmentation of blurred regions in images

Abstract

Blur detection is the separation of blurred and clear regions of an image, which is an important and challenging task in computer vision. In this work, we regard blur detection as an image segmentation problem. Inspired by the success of the U-net architecture for image segmentation, we design a Multi-Scale Dilated convolutional neural network based on U-net, which we call MSDU-net. The MSDU-net uses a group of multi-scale feature extractors with dilated convolutions to extract texture information at different scales. The U-shape architecture of the MSDU-net fuses the different-scale texture features and generates a semantic feature which allows us to achieve better results on the blur detection task. We show that using the MSDU-net we are able to outperform other state of the art blur detection methods on two publicly available benchmarks.