Vector Quantization to Visualize the Detection Process

Kaiyu Suzuki, Tomofumi Matsuzawa, Munehiro Takimoto, Yasushi Kambayashi

2021

Abstract

One of the most important tasks for drones, which are in the spotlight for assisting evacuees of natural disasters, is to automatically make decisions based on images captured by on-board cameras and provide evacuees with useful information, such as evacuation guidance. In order to make decision automatically from the aforementioned images, deep learning is the most suitable and powerful method. Although deep learning exhibits high performance, presenting the rationale for decisions is a challenge. Even though several existing decision making methods visualize and point out which part of the image they have considered intensively, they are insufficient for situations that require urgent and accurate judgments. When we look for basis for the decisions, we need to know not only WHERE to detect but also HOW to detect. This study aims to insert vector quantization (VQ) into the intermediate layer as a first step in order to show HOW to detect for deep learning in image-based tasks. We propose a method that suppresses accuracy loss while holding interpretability by applying VQ to the classification problem. The applications of the Sinkhorn–Knopp algorithm, constant embedding space and gradient penalty in this study allow us to introduce VQ with high interpretability. These techniques should help us apply the proposed method to real-world tasks where the properties of datasets are unknown.

Paper Citation

in Harvard Style

Suzuki K., Matsuzawa T., Takimoto M. and Kambayashi Y. (2021). Vector Quantization to Visualize the Detection Process.In Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 1: SDMIS, ISBN 978-989-758-484-8, pages 553-561. DOI: 10.5220/0010426005530561

in Bibtex Style

@conference{sdmis21,
author={Kaiyu Suzuki and Tomofumi Matsuzawa and Munehiro Takimoto and Yasushi Kambayashi},
title={Vector Quantization to Visualize the Detection Process},
booktitle={Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 1: SDMIS,},
year={2021},
pages={553-561},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010426005530561},
isbn={978-989-758-484-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 13th International Conference on Agents and Artificial Intelligence - Volume 1: SDMIS,
TI - Vector Quantization to Visualize the Detection Process
SN - 978-989-758-484-8
AU - Suzuki K.
AU - Matsuzawa T.
AU - Takimoto M.
AU - Kambayashi Y.
PY - 2021
SP - 553
EP - 561
DO - 10.5220/0010426005530561