Performance Evaluation of EfficientNet Variants for Image Retrieval Tasks

E. M. Martey, O. Appiah, S. K. Akpatsa, E. Antwi-Bekoe

Abstract


Content-Based Image Retrieval (CBIR) is a computational technique focused on retrieving images from a database based on their visual content. In CBIR systems, the primary focus is analysing visual features within images, enabling the matching and retrieval of relevant images. Nevertheless, deploying deep learning models for CBIR feature extraction, issues of trade-off between computational time and space complexity are of great concern. This paper explores EfficientNet, a relatively recent deep learning model famous for its computational cost efficiency and accuracy to improve the efficiency of CBIR systems. Further, the various EfficientNet model variants are investigated and analysed to assess their performance in CBIR tasks. The evaluation encompasses eight architectures within the EfficientNet family, namely EfficientNet-B0 through EfficientNet-B7, leveraging the CorelDB80 dataset as a benchmark for CBIR. The experimental results underscore the suitability of the EfficientNet family for CBIR applications, achieving a mean precision of up to 89.03%. Notably, the EfficientNet-B7 architecture consistently outperforms other variants regarding precision across different categories within the dataset. These findings provide valuable insights into the performance nuances of various EfficientNet architectures for CBIR developers, emphasizing the crucial consideration of precision in the model selection process.


Keywords


Content-Based Image Retrieval, EfficientNet, Feature Extraction, Similarity Measure, CorelDB80

Full Text:

PDF

References


Abdelrahman, A. and Viriri, S., (2023), “EfficientNet Family U-Net Models for Deep Learning Semantic Segmentation of Kidney Tumors on CT Images”. Frontiers in Computer Science, Vol. 5, pp 1235622.

Aggarwal, S., Vaid, A., Kaushik, P., Goel, A. and Kamboj, A., (2022), “Performance Evaluation of EfficientNet Model Towards Malaria Parasite Detection in Segmented Blood Cells from Thin-Blood Smear Images”. In Artificial Intelligence on Medical Data: Proceedings of International Symposium, ISCMM 2021, Singapore, pp. 137-157.

Alzubaidi, M.A., (2017), “A new strategy for bridging the semantic gap in image retrieval”. International Journal of Computational Science and Engineering, Vol. 14, No. 1, pp.27–43.

Barz, B. and Denzler, J., (2021), “Content-Based Image Retrieval and The Semantic Gap in the Deep Learning Era”. In: Del Bimbo, A., et al. Pattern Recognition, Lecture Notes in Computer Science, Vol 12662, pp. 245–260.

Chu, K. and Liu, G.H., (2020), “Image Retrieval Based on a Multi-Integration Features Model. Mathematical problems in engineering, pp.1–10.

Gherbi, T., Zeggari, A., Seghir, Z.A. and Hachouf, F., (2023), “Entropy-Guided Assessment of Image Retrieval Systems: Advancing Grouped Precision as an Evaluation Measure for Relevant Retrievability”. Informatica, Vol. 47, No. 7.

Hoang, V.T. and Jo, K.H., (2021), “Practical Analysis on the Architecture of EfficientNet”. In 2021 14th International Conference on Human System Interaction (HSI), Gdańsk, Poland, pp. 1–4. IEEE.

Jagtap, J. and Bhosle, N., (2021), “A comprehensive survey on the reduction of the semantic gap in content-based image retrieval”. International Journal of Applied Pattern Recognition, Vol. 6, No. 3, pp.254–271.

Li, J., Allinson, N., Tao, D. and Li, X., (2006), “Multitraining support vector machine for image retrieval”. IEEE Transactions on Image Processing, Vol. 15, No. 11, pp 3597-3601.

Martey, E.M., Lei, H., Li, X. and Appiah, O., (2021), “Effective Image Representation using Double Colour Histograms for Content-Based Image Retrieval”. Informatica, Vol. 45, No. 7.

Martey, E.M., Lei, H., Li, X. and Appiah, O., (2021), Image Representation Using Stacked Colour Histogram. Algorithms, Vol. 14, No. 8, pp 228.

Martey, E.M., Lei, H., Li, X., Appiah, O. and Awarayi, N.S., (2021). “Evaluation of RGB Quantization Schemes on Histogram-Based Content-Based Image Retrieval”. In Artificial Intelligence and Security: 7th International Conference, ICAIS 2021, Dublin, Ireland, (pp. 736-747). Springer International Publishing.

Mensah, M.E., Li, X., Lei, H., Obed, A. and Bombie, N.C., (2020), “Improving the Performance Of Colour-Histogram-Based CBIR using Bin Matching for Similarity Measure”. In Artificial Intelligence and Security: 6th International Conference, ICAIS 2020, Hohhot, China, pp. 586–596.

Müftüoğlu, Z., Kizrak, M.A. and Yildlnm, T., (2020), “Differential Privacy Practice on Diagnosis of COVID-19 Radiology Imaging using EfficientNet”. In 2020 International Conference on Innovations In Intelligent Systems and Applications, Novi-Sad, Serbia, pp. 1–6

Nayak, D.R., Padhy, N., Mallick, P.K., Zymbler, M. and Kumar, S., (2022), “Brain Tumour Classification using Dense EfficientNet”. Axioms, Vol 11, No. 1, pp 34.

Popescu, A., Ginsca, A. and Le-Borgne, H., (2017), “Scale-free Content Based Image Retrieval (Or Nearly So)”. In Proceedings of the IEEE International Conference on Computer Vision Workshops (pp. 280–288).

Rengel, J., Santos, M. and Pandit, R., (2022), “EfficientNet Architecture Family Analysis on Railway Track Defects”. In: Yin, H., Camacho, D., Tino, P. (eds) Intelligent Data Engineering and Automated Learning, Lecture Note in Computer Science, Vol 13756, pp. 474–481. doi: https://doi.org/ 10.1007/978-3-031-21753-1_46.

Sadia, R.T., Ahamed, M.A. and Hossain, M.A., (2021). Multiple Weather Scene Detection Utilizing the EfficientNet Family. In 2021 International Conference on Computer, Communication, Chemical, Materials and Electronic Engineering (IC4ME2), Rajshahi, Bangladesh, pp. 1-4.

Tan, M. and Le, Q., (2019), “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks”. In International conference on machine learning (pp. 6105–6114). PMLR.

Wang, W., Jiao, P., Liu, H., Ma, X. and Shang, Z., (2022), “Two-stage content-based image retrieval using sparse representation and feature fusion”. Multimedia Tools and Applications, Vol. 81, No. 12, pp.16621–16644.

Yang, J. and Zhu, S.J., (2012), “Narrowing Semantic Gap in Content-Based Image Retrieval”. In 2012 International Conference on Computer Distributed Control and Intelligent Environmental Monitoring, Zhangjiajie, China, pp. 433–438.


Refbacks

  • There are currently no refbacks.