Paper Abstract and Keywords |
Presentation |
2024-03-14 10:00
Extraction of Traffic Accident High-Risk Areas Using Deep Learning of Map Images and Grad-CAM Kaito Arase, Tsuyoshi Migita, Norikazu Takahashi (Okayama Univ.) MSS2023-86 NLP2023-138 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
An attempt has been made to predict the traffic accident risk of each map tile image at zoom level 17 using a Convolutional Neural Network (CNN) trained on map tile images labeled ``high-risk'' or ``low-risk'' based on data on traffic accidents occurred in Okayama Prefecture over the past decade. As a result, it was observed that a high prediction accuracy of about 90% was achieved and high-risk areas could be extracted in each map tile image by visualizing the decision making process of the CNN. On the other hand, it was also found that almost all map tile images were predicted to be high-risk in regions where traffic accidents are frequent, such as the central Okayama City, and that the results of Grad-CAM visualization were difficult to apply to traffic accident deterrence activities. In this study, we propose a new method for extracting high-risk areas for traffic accidents that solves the problems of the conventional method, and evaluate its effectiveness experimentally. The proposed method first labels each map tile image at zoom level 18 according to the number of traffic accidents compared to the surrounding map tile images, and then trains ResNet18, a representative CNN model, on labeled map tile images. Next, the proposed method predicts the traffic accident risk of each map tile image using the trained model. If an image is predicted to be high-risk, the proposed method identifies the risk of each of the 64 blocks obtained by dividing the image into eight sections in the vertical and horizontal directions, respectively, using the result of Grad-CAM visualization. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
traffic accident data / OpenStreetMap / convolutional neural network / Grad-CAM / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 123, no. 428, NLP2023-138, pp. 71-76, March 2024. |
Paper # |
NLP2023-138 |
Date of Issue |
2024-03-06 (MSS, NLP) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
MSS2023-86 NLP2023-138 |
Conference Information |
Committee |
NLP MSS |
Conference Date |
2024-03-13 - 2024-03-14 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Kikai-Shinko-Kaikan Bldg. |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
MSS, NLP, etc. |
Paper Information |
Registration To |
NLP |
Conference Code |
2024-03-NLP-MSS |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Extraction of Traffic Accident High-Risk Areas Using Deep Learning of Map Images and Grad-CAM |
Sub Title (in English) |
|
Keyword(1) |
traffic accident data |
Keyword(2) |
OpenStreetMap |
Keyword(3) |
convolutional neural network |
Keyword(4) |
Grad-CAM |
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Kaito Arase |
1st Author's Affiliation |
Okayama University (Okayama Univ.) |
2nd Author's Name |
Tsuyoshi Migita |
2nd Author's Affiliation |
Okayama University (Okayama Univ.) |
3rd Author's Name |
Norikazu Takahashi |
3rd Author's Affiliation |
Okayama University (Okayama Univ.) |
4th Author's Name |
|
4th Author's Affiliation |
() |
5th Author's Name |
|
5th Author's Affiliation |
() |
6th Author's Name |
|
6th Author's Affiliation |
() |
7th Author's Name |
|
7th Author's Affiliation |
() |
8th Author's Name |
|
8th Author's Affiliation |
() |
9th Author's Name |
|
9th Author's Affiliation |
() |
10th Author's Name |
|
10th Author's Affiliation |
() |
11th Author's Name |
|
11th Author's Affiliation |
() |
12th Author's Name |
|
12th Author's Affiliation |
() |
13th Author's Name |
|
13th Author's Affiliation |
() |
14th Author's Name |
|
14th Author's Affiliation |
() |
15th Author's Name |
|
15th Author's Affiliation |
() |
16th Author's Name |
|
16th Author's Affiliation |
() |
17th Author's Name |
|
17th Author's Affiliation |
() |
18th Author's Name |
|
18th Author's Affiliation |
() |
19th Author's Name |
|
19th Author's Affiliation |
() |
20th Author's Name |
|
20th Author's Affiliation |
() |
Speaker |
Author-1 |
Date Time |
2024-03-14 10:00:00 |
Presentation Time |
25 minutes |
Registration for |
NLP |
Paper # |
MSS2023-86, NLP2023-138 |
Volume (vol) |
vol.123 |
Number (no) |
no.427(MSS), no.428(NLP) |
Page |
pp.71-76 |
#Pages |
6 |
Date of Issue |
2024-03-06 (MSS, NLP) |
|