Flood video segmentation on remotely sensed UAV using improved Efficient Neural Network

Naili Suri Inthizami, M. Anwar Ma'sum, Machmud R. Alhamidi, Ahmad Gamal, Ronni Ardhianto, Kurnianingsih, Wisnu Jatmiko

Research output: Contribution to journalArticlepeer-review

Abstract

Semantic segmentation can be used to analyze the video data taken by UAV in the flood monitoring system. An accurate analysis can help rescue teams to assess and mitigate flood disasters. This paper proposed an improved Efficient Neural Network architecture to segment the UAV video of flood disaster. The proposed method consists of atrous separable convolution as the encoder and depth-wise separable convolution as the decoder. The experimental results reveal that the proposed method outperforms Efficient Neural Networks’ other architecture and gives the highest frame per second.

Original languageEnglish
JournalICT Express
DOIs
Publication statusAccepted/In press - 2022

Keywords

  • Atrous separable convolution
  • Depth-wise separable convolution
  • Efficient Neural Network
  • Semantic segmentation

Fingerprint

Dive into the research topics of 'Flood video segmentation on remotely sensed UAV using improved Efficient Neural Network'. Together they form a unique fingerprint.

Cite this