Vehicle Type Classification with Small Dataset and Transfer Learning Techniques




small dataset, Deep Learning, Transfer Learning, Knowledge Distillation


This study delves into the application of deep learning training techniques using a restricted dataset, encompassing around 400 vehicle images sourced from Kaggle. Faced with the challenges of limited data, the impracticality of training models from scratch becomes apparent, advocating instead for the utilization of pre-trained models with pre-trained weights. The investigation considers three prominent models—EfficientNetB0, ResNetB0, and MobileNetV2—with EfficientNetB0 emerging as the most proficient choice. Employing the gradually unfreeze layer technique over a specified number of epochs, EfficientNetB0 exhibits remarkable accuracy, reaching 99.5% on the training dataset and 97% on the validation dataset. In contrast, training models from scratch results in notably lower accuracy. In this context, knowledge distillation proves pivotal, overcoming this limitation and significantly improving accuracy from 29.5% in training and 20.5% in validation to 54% and 45%, respectively. This study uniquely contributes by exploring transfer learning with gradually unfreeze layers and elucidates the potential of knowledge distillation. It highlights their effectiveness in robustly enhancing model performance under data scarcity, thus addressing challenges associated with training deep learning models on limited datasets. The findings underscore the practical significance of these techniques in achieving superior results when confronted with data constraints in real-world scenarios


Download data is not yet available.

Author Biographies

Khanh-Ly Can, Phenikaa University

Phenikaa Research and Technology Institute, Hanoi, Vietnam

Hieu Dao To, Phenikaa University

Phenikaa Research and Technology Institute, Hanoi, Vietnam

Vietnam Academy of Science and Technology, Hanoi, Vietnam

Hoang-Dieu Vu, Phenikaa University

Phenikaa Research and Technology Institute, Hanoi, Vietnam

Vietnam Academy of Science and Technology, Hanoi, Vietnam


Keiron O’Shea and Ryan Nash. An introduction to convolutional neural networks. arXiv preprint arXiv:1511.08458, 2015.

Stephen Karungaru, Lyu Dongyang, and Kenji Terada. Vehicle detection and type classification based on cnn-svm. International Journal of Machine Learning and Computing, 11:304–310, 08 2021. doi:10.18178/ijmlc.2021.11.4.1052. DOI:

Hicham Bensedik, Ahmed Azough, and Meknasssi Mohammed. Vehicle type classification using convolutional neural network. pages 313–316, 10 2018. doi:10.1109/CIST.2018.8596500. DOI:

Xinchen Wang, Weiwei Zhang, Xuncheng Wu, Lingyun Xiao, Yubin Qian, and Zhi Fang. Real-time vehicle type classification with deep convolutional neural networks. Journal of Real-Time Image Processing, 16:1–10, 02 2019. doi:10.1007/s11554-017-0712-5. DOI:

Shuo Feng, Huiyu Zhou, and H.B. Dong. Using deep neural network with small dataset to predict material defects. Materials Design, 162, 11 2018. doi:10.1016/j.matdes.2018.11.060. DOI:

Miguel Romero, Yannet Interian, Timothy Solberg, and Gilmer Valdes. Training deep learning models with small datasets, 12 2019.

Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 248–255, 2009. doi:10.1109/CVPR.2009.5206848. DOI:

Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, and Liang-Chieh Chen. Mobilenetv2: Inverted residuals and linear bottlenecks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4510–4520, 2018. DOI:

Andrew G Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, and Hartwig Adam. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861, 2017.

Antonios Tragoudaras, Pavlos Stoikos, Konstantinos Fanaras, Athanasios Tziouvaras, George Floros, Georgios Dimitriou, Kostas Kolomvatsos, and Georgios Stamoulis. Design space exploration of a sparse mobilenetv2 using high-level synthesis and sparse

matrix techniques on fpgas. Sensors, 22:4318, 06 2022. doi:10.3390/s22124318. DOI:

Mingxing Tan and Quoc V. Le. Efficientnet: Rethinking model scaling for convolutional neural networks. arXiv, (1905.11946), 2020.

Chao Su and Wenjun Wang. Concrete cracks detection using convolutional neuralnetwork based on transfer learning. Mathematical Problems in Engineering, 2020:1– 10, 10 2020. doi:10.1155/2020/7240129. DOI:

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016.

Ali Bhatti, Muhammad Umer, Syed Adil, Mansoor Ebrahim, Daniyal Nawaz, and Faizan Ahmed. Explicit content detection system: An approach towards a safe and ethical environment. Applied Computational Intelligence and Soft Computing, 2018, 07 2018. doi:10.1155/2018/1463546. DOI:

Geoffrey Hinton, Oriol Vinyals, and Jeff Dean. Distilling the knowledge in a neural network. arXiv, (1503.02531), 2015.

Jianping Gou, Baosheng Yu, Stephen J. Maybank, and Dacheng Tao. Knowledge distillation: A survey. International Journal of Computer Vision, 129(6):1789–1819, March 2021. ISSN 1573-1405. doi:10.1007/s11263-021-01453-z. URL DOI:




How to Cite

Pham, Q.-T., Pham, D.-D., Can, K.-L., Dao To, H., & Vu, H.-D. (2024). Vehicle Type Classification with Small Dataset and Transfer Learning Techniques. EAI Endorsed Transactions on Industrial Networks and Intelligent Systems, 11(2), e2.

Funding data