Unsupervised Domain Adaptation for Environmental Recognition in Crane Operations

Keigo Watanabe, Maierdan Maimaitimin, Yuta Takashima, Isaku Nagai

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In loading work with a helicopter or a crane, the body itself recognizes an environment, and in order to develop a system that judges whether the commands and operations by human are safe, it needs to realize semantic recognition of the environment, distance recognition to an obstacle, and classification and pursuit of moving objects with high precision using sensors. A method of realizing semantic segmentation, depth estimation and optical flow simultaneously from a camera image had been proposed with a multitasking DNN that took account of the posture and speed of a drone considering a loading work in the air, and it was proved to be useful in a simulated environment. Note however that the model learned in the simulation environment is not a thing suitable for the environmental recognition in a real world. Therefore, this paper aims to develop an environmental recognition system that can be used in the real world by conducting a domain adaptation with adversarial learning. The usefulness of the domain adaptation technique in the proposed multitasking DNN is verified by carrying out environmental recognition of the actual image acquired from the boom tip of a crane.

Original languageEnglish
Title of host publicationProceedings of 2022 SICE International Symposium on Control Systems, SICE ISCS 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages80-86
Number of pages7
ISBN (Electronic)9784907764746
DOIs
Publication statusPublished - 2022
Event2022 SICE International Symposium on Control Systems, SICE ISCS 2022 - Virtual, Online, Japan
Duration: Mar 8 2022Mar 10 2022

Publication series

NameProceedings of 2022 SICE International Symposium on Control Systems, SICE ISCS 2022

Conference

Conference2022 SICE International Symposium on Control Systems, SICE ISCS 2022
Country/TerritoryJapan
CityVirtual, Online
Period3/8/223/10/22

Keywords

  • Deep Neural Networks
  • Multi-task
  • Transfer Learning
  • Unsupervised Domain Adaptation

ASJC Scopus subject areas

  • Artificial Intelligence
  • Control and Optimization
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Unsupervised Domain Adaptation for Environmental Recognition in Crane Operations'. Together they form a unique fingerprint.

Cite this