Saved in:
Bibliographic Details
Main Authors: Kiyokawa, Takuya, Shirakura, Naoki, Katayama, Hiroki, Tomochika, Keita, Takamatsu, Jun
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2304.04901
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909379190587392
author Kiyokawa, Takuya
Shirakura, Naoki
Katayama, Hiroki
Tomochika, Keita
Takamatsu, Jun
author_facet Kiyokawa, Takuya
Shirakura, Naoki
Katayama, Hiroki
Tomochika, Keita
Takamatsu, Jun
contents Training deep-learning-based vision systems require the manual annotation of a significant number of images. Such manual annotation is highly time-consuming and labor-intensive. Although previous studies have attempted to eliminate the effort required for annotation, the effort required for image collection was retained. To address this, we propose a human-in-the-loop dataset collection method that uses a web application. To counterbalance the workload and performance by encouraging the collection of multi-view object image datasets in an enjoyable manner, thereby amplifying motivation, we propose three types of online visual feedback features to track the progress of the collection status. Our experiments thoroughly investigated the impact of each feature on collection performance and quality of operation. The results suggested the feasibility of annotation and object detection.
format Preprint
id arxiv_https___arxiv_org_abs_2304_04901
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Efficiently Collecting Training Dataset for 2D Object Detection by Online Visual Feedback
Kiyokawa, Takuya
Shirakura, Naoki
Katayama, Hiroki
Tomochika, Keita
Takamatsu, Jun
Computer Vision and Pattern Recognition
Multimedia
Training deep-learning-based vision systems require the manual annotation of a significant number of images. Such manual annotation is highly time-consuming and labor-intensive. Although previous studies have attempted to eliminate the effort required for annotation, the effort required for image collection was retained. To address this, we propose a human-in-the-loop dataset collection method that uses a web application. To counterbalance the workload and performance by encouraging the collection of multi-view object image datasets in an enjoyable manner, thereby amplifying motivation, we propose three types of online visual feedback features to track the progress of the collection status. Our experiments thoroughly investigated the impact of each feature on collection performance and quality of operation. The results suggested the feasibility of annotation and object detection.
title Efficiently Collecting Training Dataset for 2D Object Detection by Online Visual Feedback
topic Computer Vision and Pattern Recognition
Multimedia
url https://arxiv.org/abs/2304.04901