Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN
Convolutional Neural Networks (CNN) achieves perfection in traffic sign identification with enough annotated training data. The dataset determines the quality of the complete visual system based on CNN. Unfortunately, databases for traffic signs from the majority of the world’s nations ar...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9471877/ |
| _version_ | 1818355393679589376 |
|---|---|
| author | Christine Dewi Rung-Ching Chen Yan-Ting Liu Xiaoyi Jiang Kristoko Dwi Hartomo |
| author_facet | Christine Dewi Rung-Ching Chen Yan-Ting Liu Xiaoyi Jiang Kristoko Dwi Hartomo |
| author_sort | Christine Dewi |
| collection | DOAJ |
| description | Convolutional Neural Networks (CNN) achieves perfection in traffic sign identification with enough annotated training data. The dataset determines the quality of the complete visual system based on CNN. Unfortunately, databases for traffic signs from the majority of the world’s nations are few. In this scenario, Generative Adversarial Networks (GAN) may be employed to produce more realistic and varied training pictures to supplement the actual arrangement of images. The purpose of this research is to describe how the quality of synthetic pictures created by DCGAN, LSGAN, and WGAN is determined. Our work combines synthetic images with original images to enhance datasets and verify the effectiveness of synthetic datasets. We use different numbers and sizes of images for training. Likewise, the Structural Similarity Index (SSIM) and Mean Square Error (MSE) were employed to assess picture quality. Our study quantifies the SSIM difference between the synthetic and actual images. When additional images are used for training, the synthetic image exhibits a high degree of resemblance to the genuine image. The highest SSIM value was achieved when using 200 total images as input and <inline-formula> <tex-math notation="LaTeX">$32\times 32$ </tex-math></inline-formula> image size. Further, we augment the original picture dataset with synthetic pictures and compare the original image model to the synthesis image model. For this experiment, we are using the latest iterations of Yolo, Yolo V3, and Yolo V4. After mixing the real image with the synthesized image produced by LSGAN, the recognition performance has been improved, achieving an accuracy of 84.9% on Yolo V3 and an accuracy of 89.33% on Yolo V4. |
| first_indexed | 2024-12-13T19:40:37Z |
| format | Article |
| id | doaj.art-6e06fd041ba3447ba0f362c016134c31 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-13T19:40:37Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-6e06fd041ba3447ba0f362c016134c312022-12-21T23:33:41ZengIEEEIEEE Access2169-35362021-01-019972289724210.1109/ACCESS.2021.30942019471877Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GANChristine Dewi0https://orcid.org/0000-0002-1284-234XRung-Ching Chen1https://orcid.org/0000-0001-7621-1988Yan-Ting Liu2Xiaoyi Jiang3https://orcid.org/0000-0001-7678-9528Kristoko Dwi Hartomo4https://orcid.org/0000-0003-0237-851XDepartment of Information Management, Chaoyang University of Technology, Taichung, TaiwanDepartment of Information Management, Chaoyang University of Technology, Taichung, TaiwanDepartment of Information Management, Chaoyang University of Technology, Taichung, TaiwanDepartment of Mathematics and Computer Science, University of Münster, Münster, GermanyFaculty of Information Technology, Satya Wacana Christian University, Central Java, Salatiga, IndonesiaConvolutional Neural Networks (CNN) achieves perfection in traffic sign identification with enough annotated training data. The dataset determines the quality of the complete visual system based on CNN. Unfortunately, databases for traffic signs from the majority of the world’s nations are few. In this scenario, Generative Adversarial Networks (GAN) may be employed to produce more realistic and varied training pictures to supplement the actual arrangement of images. The purpose of this research is to describe how the quality of synthetic pictures created by DCGAN, LSGAN, and WGAN is determined. Our work combines synthetic images with original images to enhance datasets and verify the effectiveness of synthetic datasets. We use different numbers and sizes of images for training. Likewise, the Structural Similarity Index (SSIM) and Mean Square Error (MSE) were employed to assess picture quality. Our study quantifies the SSIM difference between the synthetic and actual images. When additional images are used for training, the synthetic image exhibits a high degree of resemblance to the genuine image. The highest SSIM value was achieved when using 200 total images as input and <inline-formula> <tex-math notation="LaTeX">$32\times 32$ </tex-math></inline-formula> image size. Further, we augment the original picture dataset with synthetic pictures and compare the original image model to the synthesis image model. For this experiment, we are using the latest iterations of Yolo, Yolo V3, and Yolo V4. After mixing the real image with the synthesized image produced by LSGAN, the recognition performance has been improved, achieving an accuracy of 84.9% on Yolo V3 and an accuracy of 89.33% on Yolo V4.https://ieeexplore.ieee.org/document/9471877/DCGANLSGANsynthetic imagestraffic signWGANYolo V3 |
| spellingShingle | Christine Dewi Rung-Ching Chen Yan-Ting Liu Xiaoyi Jiang Kristoko Dwi Hartomo Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN IEEE Access DCGAN LSGAN synthetic images traffic sign WGAN Yolo V3 |
| title | Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN |
| title_full | Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN |
| title_fullStr | Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN |
| title_full_unstemmed | Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN |
| title_short | Yolo V4 for Advanced Traffic Sign Recognition With Synthetic Training Data Generated by Various GAN |
| title_sort | yolo v4 for advanced traffic sign recognition with synthetic training data generated by various gan |
| topic | DCGAN LSGAN synthetic images traffic sign WGAN Yolo V3 |
| url | https://ieeexplore.ieee.org/document/9471877/ |
| work_keys_str_mv | AT christinedewi yolov4foradvancedtrafficsignrecognitionwithsynthetictrainingdatageneratedbyvariousgan AT rungchingchen yolov4foradvancedtrafficsignrecognitionwithsynthetictrainingdatageneratedbyvariousgan AT yantingliu yolov4foradvancedtrafficsignrecognitionwithsynthetictrainingdatageneratedbyvariousgan AT xiaoyijiang yolov4foradvancedtrafficsignrecognitionwithsynthetictrainingdatageneratedbyvariousgan AT kristokodwihartomo yolov4foradvancedtrafficsignrecognitionwithsynthetictrainingdatageneratedbyvariousgan |