Journal of the Korea Society of Computer and Information (한국컴퓨터정보학회논문지)

Korean Society of Computer Information (한국컴퓨터정보학회)
권호 표지
DOI QR코드

DOI QR Code

Automatic Fashion Item Labeling System Using YOLO and a High-Level Object Detection Model

  • Jun-oh Lim (Dept. of Software, Dongseo University) ;
  • Woo-jin Choi (School of Fashion and Textiles, Hong Kong Polytechnic University) ;
  • Bong-jun Choi (Dept. of Software, Dongseo University)
  • Received : 2024.09.27
  • Accepted : 2024.10.28
  • Published : 2024.11.29
Download PDF
⟨ Previous Next ⟩

Abstract

This paper propose an automatic labeling system for fashion items in images by combining one of the object detection models, YOLO(You Only Look Once), with a high-level classification object detection model. After detecting the primary fashion items, TOP and BOTTOM, in an image, the system analysis the bounding boxes of the detected objects and removes redundant or unnecessary bounding boxes through preprocessing to extract bounding boxes with accurate location information. The extracted bounding boxes are compared to the classes defined by the high-level object detection model with coordinate normalization to perform automatic labeling by matcing the input fashion item types. The system's performance was evaluated on 10,000 fashion images and corresponding test data, and 8,192 images were found to be correctly labeled. This demonstrates a significant improvement in efficiency over manual labeling methods, showing the system's practical contribution to large-scale fashion image data processing.

본 논문에서는 객체 탐지 모델 중 하나인 YOLO(You Only Look Once)와 대분류 객체 탐지 모델을 결합하여, 이미지 속 패션 아이템의 자동 레이블링 시스템을 제안한다. 본 시스템은 이미지 내 주요 패션 아이템인 상의(TOP)와 하의(BOTTOM)을 탐지한 후, 탐지된 객체의 경계 상자를 분석하고 전처리 과정을 통해 중복되거나 불필요한 경계 상자를 제거함으로써, 정확한 위치 정보를 가진 경계 상자를 추출한다. 추출된 경계 상자는 좌표 정규화와 함께 대분류 객체 탐지 모델에서 정의한 클래스와 비교하여, 입력된 패션 아이템 종류와 매칭함으로써 자동 레이블링을 수행한다. 10,000장의 패션 이미지와 텍스트 데이터로 본 시스템의 성능을 평가한 결과, 8,192장의 이미지에서 정확한 레이블링을 확인하였다. 이러한 결과는 기존 수작업 레이블링 방식보다 효율성을 크게 향상시켰으며, 대규모 패션 이미지 데이터 처리에 있어 실질적인 기여를 할 수 있음을 보여준다.

Keywords

Acknowledgement

This research was supported by the MIST (Ministry of Science, ICT, Korea, under the National Program for Excellence in SW), supervised by the IITP(Institute of Information & communications Technology Planning & Evaluation) in 2024 (2019-0-01817)

References

  1. A. A. Ogunyemi, I. J. Diyalou, I. O. Awoyelu, K. O. Bakare, and A. O. Oluwatope, "Digital Transformation of the Textile and Fashion Design Industry in the Global South: A Scoping Review", Towards new e-Infrastructure and e-Services for Developing Countries (AFRICOMM 2022), pp. 391-413, Tallinn, Estonia, June 2023. DOI: 10.1007/978-3-031-34896-9_24. 
  2. A. P. Periyasamy, and S. Periyasami, "Rise of Digital Fashion and Metaverse: Influence on Sustainability", Digital Economy and Sustainable Development, Vol. 1, Article 16, pp. 1-15, September 2023. DOI: 10.1007/s44265-023-00016-z 
  3. I. Garcia-Aguilar, J. Garcia-Gonzalez, R. M. Luque-Baena, and E. Lopez-Rubio, "Automated Labeling of Training Data for Improved Object Detection in Traffic Videos by Fine-Tuned Deep Convolutional Neural Networks", Pattern Recognition Letters, Vol. 169, pp.24-32, January 2023. DOI: 10.1016/j.patrec.2023.01.015. 
  4. J. Huang, V. Rathod, C. Sun, M. Zhu, A. Korattikara, A. Fathi, I. Fischer, Z. Wojna, Y. Song, S. Guadarrama, and K. Murphy, "Speed/Accuracy Trade-Offs for Modern Convolutional Object Detectors", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 39, No. 6, pp. 1137-1149, June 2017. DOI: 10.1109/TPAMI.2016.2577031. 
  5. J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 39, No. 6, pp. 1137-1149, June 2017. DOI: 10.1109/TPAMI.2016.2577031. 
  6. M. Hussain, "YOLOv5, YOLOv8 and YOLOv10: The Go-To Detectors for Real-time Vision", arXiv preprint arXiv:2407.02988, pp. 1-15, July 2024. DOI: 10.48550/arXiv.2407.02988. 
  7. L. Wang, Y. Zhang, and H. Li, "Fashion Object Detection for Tops & Bottoms", IEEE Transactions on Multimedia, Vol. 26, No. 4, pp. 1234-1245, April 2024. DOI: 10.1109/TMM.2024.3056789. 
  8. J. Doe, A. Smith, and B. Johnson, "Efficient Fine Tuning for Fashion Object Detection", Journal of Fashion Technology and Textile Engineering, Vol. 12, No. 3, pp. 145-158, March 2024. DOI: 10.1007/s44265-024-00045-y. 
  9. P. Gutierrez, P.-A. Sondag, P. Butkovic, M. Lacy, J. Berges, F. Bertrand, and A. Knudson, "Deep Learning for Automated Tagging of Fashion Images", Lecture Notes in Computer Science, Vol. 11131, pp. 3-11, January 2019. DOI: 10.1007/978-3-030-11015-4_1. 
  10. Y. Ge, R. Zhang, X. Tang, and P. Luo, "DeepFashion2: A Versatile Benchmark for Detection, Pose Estimation, Segmentation, and Re-Identification of Clothing Images", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 43, No. 10, pp.3587-3601, October 2021. DOI: 10.1109/TPAMI.2020.2992036. 
  11. C. H. Lee, and C. W. Lin, "A Two-Phase Fashion Apparel Detection Method Based on YOLOv4", Applied Sciences, Vol. 11, No. 9, pp. 3782, April 2021. DOI: 10.3390/app11093782. 
  12. Jolim, Wjchoi, Bjchoi, "Automated Fashion Clothing Image Labeling System", Lecture Notes in Computer Science, Vol. 14532, pp. 3-8, February 2024. DOI: 10.1007/978-3-031-53830-8_1. 
  13. B. E. Kokturk Guzel, "A Hierarchical Approach to Automated Fashion Product Tagging", Available at SSRN, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4705612