V. V. Moskalenko, A. S. Moskalenko, A. G. Korobov, M. O. Zaretsky


Context. Lightweight model and effective training algorithm of on-board object detection system for a compact drone are developed. The object of research is the process of small object detection on aerial images under computational resource constraint and uncertainty caused by small amount of labeled training data. The subject of the research are the model and training algorithm for detecting small objects on aerial imagery.
Objective. Goal of the research is developing efficient model and training algorithm of object detection system for a compact aerial drone under conditions of restricted computing resources and the limited volume of the labeled training set.
Methods. The four-stage training algorithm of the object detector is proposed. At the first stage, selecting the type of deep convolutional neural network and the number of low-level layers that is pre-trained on the ImageNet dataset for reusing takes place. The second stage involves unsupervised training of high-level convolutional sparse coding layers using modification of growing neural gas to automatically determine the required number of neurons and provide optimal distributions of the neurons over the data. Its application makes it possible to utilize the unlabeled training datasets for the adaptation of the high-level feature description to the domain application area. At the third stage, there are reduction of output feture map using principan component analysis and building of decision rules. In this case, output feature map is formed by concatenation of feature maps from different level of deep network using upscaling upper maps to uniform shape for each channel. This approach provides more contextual information for efficient recognition of small objects on aerial images. In order to build classifier of output feature map pixels is proposed to use boosted
information-extreme learning. Besides that, the extreme learning machine is used to build of regression model for predict bounding box of detected object. Last stage involves fine-tuning of high-level layers of deep network using simulated annealing metaheuristic algorithm in order to approximating the global optimum of complex criterion of training efficiency of detection model.
Results. The results obtained on open datasets testify to the suitability of the model and training algorithm for practical usage. The proposed training algorithm utilize 500 unlabeled and 200 labeled training samples to provide 96% correctly detection of objects on the images of the test dataset.
Conclusions. Scientific novelty of the paper is a new model and training algorithm for object detection, which enable to achieve high confidence of recognition of small objects on aerial images under computational resource constraint and limited volume of the labeled training set. Practical significance of the paper results consists in the developed model and training algorithm made it possible to reduce requirements for size of labeled training set and computation resources of on-board detection system of aerial drone in training and inference modes.


growing neural gas; convolutional neural network; boosting; objects detector; information criterion; simulated annealing algorithm; extreme learning.


Patricia N., Caputo B. Learning to Learn, from Transfer

Learning to Domain Adaptation: A Unifying Perspective,

IEEE Conference on Computer Vision and Pattern

Recognition (CVPR), Columbus, Ohio, 23–28 June, 2014 :

proceedings. Conference Publishing Services, 2014,

P. 1442–1449. DOI: 10.1109/CVPR.2014.187.

Nguyen A., Yosinski J., Clune J. Deep neural networks are

easily fooled: High confidence predictions for

unrecognizable images, IEEE Conference on Computer

Vision and Pattern Recognition, (CVPR), Boston, MA, 7–12

June, 2015 : proceedings, Conference Publishing Services,

, P. 427–436. DOI: 10.1109/CVPR.2015.7298640.

Ayumi V., Rere L. M. R., Fanany M. I., Arymurthy A. M.

Optimization of convolutional neural network using

microcanonical annealing algorithm / V. Ayumi, //

International Conference on Advanced Computer Science

and Information Systems (ICACSIS), Malang, Indonesia,

–16 Oct., 2016: proceedings. Institute of Electrical and

Electronics Engineers, 2016, pp. 506–511.

DOI: 10.1109/ICACSIS.2016.7872787.

Antipov G. Berrani S., Ruchaud N., Dugelay J. Learned vs.

Hand-Crafted Features for Pedestrian Gender Recognition,

rd ACM International Conference on Multimedia,

Brisbane, Australia, 26–30 Oct., 2015: proceedings, ACM.

New York, NY, USA, 2015, pp. 1263–1266. DOI:


Carrio A. Sampedro C., Rodriguez-Ramos A., Campoy P. A

Review of Deep Learning Methods and Applications for

Unmanned Aerial Vehicles, Journal of Sensors, 2017, Vol.

– P. 1–13. DOI:10.1155/2017/3296874.

Subbotin S. The special deep neural network for stationary

signal spectra classification, 14th International Conference

on Advanced Trends in Radioelecrtronics,

Telecommunications and Computer Engineering (TCSET).

Lviv-Slavske, Ukraine, 20–24 Feb, 2018, proceedings,

IEEE, 2018, pp.123–128.

Xu X., Ding Y., Hu S. X. Scaling for edge inference of deep

neural networks, Nature Electronics, 2018, Vol. 1(4),

pp. 216–222. DOI:10.1038/s41928-018-0059-3.

Loquercio A. Maqueda A. I., del-Blanco C. R.,

Scaramuzza D. DroNet: Learning to Fly by Driving, IEEE

Robotics and Automation Letters, 2018, Vol. 3, No. 2,

pp. 1088–1095. DOI: 10.1145/2733373.2806332.

Mathew A., Mathew J., Govind M., Mooppan A. An

Improved Transfer learning Approach for Intrusion

Detection, Procedia Computer Science, 2017, Vol. 115,

pp. 251–257. DOI: 10.1016/j.procs.2017.09.132.

Radenović F., Tolias G., Chum O. Fine-tuning CNN Image

Retrieval with No Human Annotation, IEEE Transactions

on Pattern Analysis and Machine Intelligence, 2018, Mode

of access:

DOI: 10.1109/TPAMI.2018.2846566.

Moskalenko V., Dovbysh S. , Naumenko I., Moskalenko A.,

Korobov A. Improving the effectiveness of training the onboard

object detection system for a compact unmanned

aerial vehicle, Eastern-European Journal of Enterprise

Technologies, 2018, No. 4/9 (94), pp. 19–26.

DOI: 10.15587/1729-4061.2018.139923

Vens C., Costa F. Random Forest Based Feature Induction,

IEEE 11th International Conference on Data Mining,

Vancouver, Canada, 11–14 Dec, 2011: proceedings,

pp. 744–753. DOI: 10.1109/ICDM.2011.121

Feng Q., Chen C. L. P., Chen L. Compressed auto-encoder

building block for deep learning network, 3rd International

Conference on Informative and Cybernetics for

Computational Social Systems (ICCSS), Jinzhou, Liaoning,

China, 26–29 Aug, 2016: proceedings, IEEE, 2016,

pp. 131–136. DOI: 10.1109/ICCSS.2016.7586437

Labusch K., Barth E., Martinetz T. Sparse coding neural

gas: learning of overcomplete data representations,

Neurocomputing, 2009, Vol. 72, I. 7–9, pp. 1547–1555.


Mrazova I., Kukacka M. Image Classification with Growing

Neural Networks, International Journal of Computer Theory

and Engineering, 2013, Vol. 5, No. 3, pp. 422–427.


GOST Style Citations

Copyright (c) 2019 V. V. Moskalenko, A. S. Moskalenko, A. G. Korobov, M. O. Zaretsky

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Address of the journal editorial office:
Editorial office of the journal «Radio Electronics, Computer Science, Control»,
Zaporizhzhya National Technical University, 
Zhukovskiy street, 64, Zaporizhzhya, 69063, Ukraine. 
Telephone: +38-061-769-82-96 – the Editing and Publishing Department.

The reference to the journal is obligatory in the cases of complete or partial use of its materials.