Target-based visual navigation with channel-aware network

Huichao Li, Xuemei Ren, Yongfeng Lv

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Visual navigation is major content of robot control, especially those based on target. We propose a channel-aware deep siamese actor-critic network for target-based visual navigation task. Compared with previous target-driven network, our model can obtain a joint representation of siamese network's output feature by using distance fusion method, and the approach significantly accelerates convergence of model's training. We improve the model performance to make the agent reach the goal in shorter path during the navigation by inserting a modified Squeeze-and-Excitation block in siamese layers, in which way the model can take dependencies between visual feature channels into consideration.

Original languageEnglish
Title of host publicationProceedings of 2019 IEEE 8th Data Driven Control and Learning Systems Conference, DDCLS 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages838-843
Number of pages6
ISBN (Electronic)9781728114545
DOIs
Publication statusPublished - May 2019
Event8th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2019 - Dali, China
Duration: 24 May 201927 May 2019

Publication series

NameProceedings of 2019 IEEE 8th Data Driven Control and Learning Systems Conference, DDCLS 2019

Conference

Conference8th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2019
Country/TerritoryChina
CityDali
Period24/05/1927/05/19

Keywords

  • AI2-THOR framework
  • Data fusion
  • Reinforcement learning
  • Siamese network
  • Squeeze-and-excitation

Fingerprint

Dive into the research topics of 'Target-based visual navigation with channel-aware network'. Together they form a unique fingerprint.

Cite this