![](bg6.png)
ods like iCARL or Maximal Interfered Retrieval (Re-
buffi et al., 2017; Aljundi et al., 2019) will be use-
ful. These methods can then be deployed on the Jet-
son Nano setup itself towards an on-edge solution that
continually adapts with minimal manual intervention.
Furthermore, there have been interesting advances
in the area of Unsupervised Continual Learning (Rao
et al., 2019; He and Zhu, 2021; Bertugli et al., 2020).
Such an approach could not only aid in solving the
issues discussed in this paper but can also allow for
the product using the detection model to be much
more scalable to new environments without manual
labelling. Hence, comparing the results and practi-
cality of an unsupervised method to the incremental
labelling approach is another area for further investi-
gation.
ACKNOWLEDGEMENTS
This research is supported by the Ministry of Uni-
versity and Research (MUR) as part of the PON
2014-2020 “Research and Innovation” resources –
Green/Innovation Action – DM MUR 1061/2022.
This research was also supported by the Nvidia Aca-
demic Hardware Grant Program.
REFERENCES
Alam, N.-A., Ahsan, M., Based, M. A., and Haider, J.
(2021). Intelligent system for vehicles number plate
detection and recognition using convolutional neural
networks. Technologies, 9(1):9.
Aljundi, R., Belilovsky, E., Tuytelaars, T., Charlin, L., Cac-
cia, M., Lin, M., and Page-Caccia, L. (2019). Online
continual learning with maximal interfered retrieval.
Advances in neural information processing systems,
32.
Bertugli, A., Vincenzi, S., Calderara, S., and Passerini,
A. (2020). Few-shot unsupervised continual
learning through meta-examples. arXiv preprint
arXiv:2009.08107.
Biesialska, M., Biesialska, K., and Costa-juss
`
a, M. R.
(2020). Continual lifelong learning in natural
language processing: A survey. arXiv preprint
arXiv:2012.09823.
Chen, J. and Ran, X. (2019). Deep learning with edge
computing: A review. Proceedings of the IEEE,
107(8):1655–1674.
Cossu, A., Carta, A., Lomonaco, V., and Bacciu, D. (2021).
Continual learning for recurrent neural networks: an
empirical evaluation. Neural Networks, 143:607–627.
Delange, M., Aljundi, R., Masana, M., Parisot, S., Jia,
X., Leonardis, A., Slabaugh, G., and Tuytelaars, T.
(2021). A continual learning survey: Defying forget-
ting in classification tasks. IEEE Transactions on Pat-
tern Analysis and Machine Intelligence.
Demosthenous, G. and Vassiliades, V. (2021). Contin-
ual learning on the edge with tensorflow lite. arXiv
preprint arXiv:2105.01946.
Goodfellow, I. J., Mirza, M., Xiao, D., Courville, A., and
Bengio, Y. (2013). An empirical investigation of
catastrophic forgetting in gradient-based neural net-
works. arXiv preprint arXiv:1312.6211.
He, J. and Zhu, F. (2021). Unsupervised continual learning
via pseudo labels. arXiv preprint arXiv:2104.07164.
Huang, Z., Yang, S., Zhou, M., Gong, Z., Abusorrah, A.,
Lin, C., and Huang, Z. (2021). Making accurate ob-
ject detection at the edge: review and new approach.
Artificial Intelligence Review, pages 1–30.
Jamtsho, Y., Riyamongkol, P., and Waranusast, R. (2021).
Real-time license plate detection for non-helmeted
motorcyclist using yolo. Ict Express, 7(1):104–109.
Jocher, G., Chaurasia, A., Stoken, A., Borovec, J.,
NanoCode012, Kwon, Y., TaoXie, Fang, J., imyhxy,
Michael, K., Lorna, V, A., Montes, D., Nadar, J.,
Laughing, tkianai, yxNONG, Skalski, P., Wang, Z.,
Hogan, A., Fati, C., Mammana, L., AlexWang1900,
Patel, D., Yiwei, D., You, F., Hajek, J., Diaconu, L.,
and Minh, M. T. (2022). ultralytics/yolov5: v6.1 -
TensorRT, TensorFlow Edge TPU and OpenVINO Ex-
port and Inference.
Khetarpal, K., Riemer, M., Rish, I., and Precup, D. (2020).
Towards continual reinforcement learning: A review
and perspectives. arXiv preprint arXiv:2012.13490.
Kwon, Y. D., Chauhan, J., Kumar, A., Hui, P., and Mascolo,
C. (2021). Exploring system performance of contin-
ual learning for mobile and embedded sensing appli-
cations. arXiv preprint arXiv:2110.13290.
Li, H., Ota, K., and Dong, M. (2018). Learning iot in
edge: Deep learning for the internet of things with
edge computing. IEEE network, 32(1):96–101.
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P.,
Ramanan, D., Doll
´
ar, P., and Zitnick, C. L. (2014).
Microsoft coco: Common objects in context. In Euro-
pean conference on computer vision, pages 740–755.
Springer.
Liu, X., Yang, H., Ravichandran, A., Bhotika, R., and
Soatto, S. (2020). Continual universal object detec-
tion. arXiv preprint arXiv:2002.05347.
Lomonaco, V., Pellegrini, L., Cossu, A., Carta, A., Graffi-
eti, G., Hayes, T. L., De Lange, M., Masana, M., Pom-
poni, J., Van de Ven, G. M., et al. (2021). Avalanche:
an end-to-end library for continual learning. In Pro-
ceedings of the IEEE/CVF Conference on Computer
Vision and Pattern Recognition, pages 3600–3610.
Mai, Z., Li, R., Jeong, J., Quispe, D., Kim, H., and San-
ner, S. (2022). Online continual learning in image
classification: An empirical survey. Neurocomputing,
469:28–51.
Parisi, G. I., Kemker, R., Part, J. L., Kanan, C., and
Wermter, S. (2019). Continual lifelong learning with
neural networks: A review. Neural Networks, 113:54–
71.
Using Continual Learning on Edge Devices for Cost-Effective, Efficient License Plate Detection
549