Indoor mapping using audio reflections from mobile devices

With the advancement of mobile computing technologies, localization and mapping are becoming increasingly important topics in recent years. While a lot of work is appearing in the localization domain, mapping has much limited options. Mapping, more often than not, would require high resolution imaging sensors, which would reveal privacy information. In this thesis, we explore the possibility of privacy preserving indoor mapping using inertial and acoustic sensors on mobile devices, without using cameras. We addressed a number of challenges including calibrating inertial sensor drift and reasoning from noisy acoustic data. We built an end-to-end mapping pipeline, and evaluated it in real-world environments. We demonstrated a spatial audio based navigation application on top of the mapping result. The results are promising, and furthermore, this opens up the future research direction in the intersection of mobile computing, sensing, signal processing, and machine learning.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2023

Author dc:creator

• Yang, Zhijian

Contributors dc:contributor

• Roy Choudhury, Romit
• Nahrstedt, Klara
• Vasisht, Deepak
• Shangguan, Longfei

• Localization And Mapping
• Sensing
• Mobile Computing

Statement dc:rights

• Copyright 2023 Zhijian Yang

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/122140

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F122140