Lightweight the Focus module in YOLOv5 by Dilated Convolution
Authors:
- 20 May 2022
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 111-114
Abstract
Focus is an advanced module in YOLOv5. However, no detailed work has been published to demonstrate its effectiveness so far. In this paper, we analyze the Focus module from two respects, FLOPs and receptive field, and reveal that it’s approximated to convolution layer with the same kernel size, twice the receptive field and quadruple the computational cost. Inspired by dilated convolution, we propose a new module named DFocus. DFocus is comparable in performance to Focus but without any additional computation cost. The experiments on PASCAL VOC dataset demonstrate the effectiveness of our work.Keywords
This publication has 13 references indexed in Scilit:
- M2Det: A Single-Shot Object Detector Based on Multi-Level Feature Pyramid NetworkProceedings of the AAAI Conference on Artificial Intelligence, 2019
- IR Feature Embedded BOF Indexing Method for Near-Duplicate Video RetrievalIEEE Transactions on Circuits and Systems for Video Technology, 2018
- Parallel Feature Pyramid Network for Object Detection2018
- A Dynamic Analytics Method Based on Multistage Modeling for a BOF Steelmaking ProcessIEEE Transactions on Automation Science and Engineering, 2018
- Understanding of Object Detection Based on CNN Family and YOLOJournal of Physics: Conference Series, 2018
- Feature Pyramid Networks for Object Detection2017
- Deep Residual Learning for Image Recognition2016
- Application of background oriented Schlieren (BOS) method for visualization of evaporating impinged sprayOptik, 2015
- Background-oriented schlieren (BOS) techniquesExperiments in Fluids, 2015
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009