A robotic assistant for pictogram classification in education: a proposal using dynamically quantized deep neural networks and the CIFAR-10 dataset for developing countries

Lisseth Padilla-Viñanzaca; Bryam Guachún-Guamán; Vladimir Robles-Bykbaev; Efrén Lema-Condo

doi:10.1590/1983-3652.2026.58683

Autores/as

Lisseth Padilla-Viñanzaca Universidad Politécnica Salesiana, Cátedra UNESCO Tecnologías de apoyo para la Inclusión Educativa, Cuenca, Azuay, Ecuador https://orcid.org/0009-0002-8657-5018
Bryam Guachún-Guamán Universidad Politécnica Salesiana, Cátedra UNESCO Tecnologías de apoyo para la Inclusión Educativa, Cuenca, Azuay, Ecuador https://orcid.org/0009-0000-2319-3704
Vladimir Robles-Bykbaev Universidad Politécnica Salesiana, Cátedra UNESCO Tecnologías de apoyo para la Inclusión Educativa, Cuenca, Azuay, Ecuador https://orcid.org/0000-0002-7645-8793
Efrén Lema-Condo Universidad Politécnica Salesiana, Cátedra UNESCO Tecnologías de apoyo para la Inclusión Educativa, Cuenca, Azuay, Ecuador https://orcid.org/0000-0002-6710-3683

DOI:

https://doi.org/10.1590/1983-3652.2026.58683

Palabras clave:

Primary school students, Free Educational Robotics, Computer uses in education, Artificial intelligence

Resumen

Currently, there exists a wide range of deep learning models developed for numerous tasks, ranging from automatic speech recognition to music and video generation. According to various authors, these models hold significant potential to contribute to achieving several Sustainable Development Goals (SDGs) established by the United Nations. However, in developing countries such as Ecuador, not all educational institutions -particularly those in rural areas- have access to the necessary infrastructure to implement these models in ways that enhance educational processes for children. In response to this issue, this study presents a low-cost robotic assistant that utilizes quantized deep learning networks to support the recognition of pictograms in basic general education. The proposed system was tested with a group of 52 children between the ages of 5 and 8, yielding a Cronbach's Alpha coefficient of 0.71, which suggests that the solution is promising.

Descargas

Los datos de descarga aún no están disponibles.

Referencias

BELPAEME, Tony; KENNEDY, James; RAMACHANDRAN, Aditi; SCASSELLATI, Brian; TANAKA, Fumihide. Social robots for education: A review. Science robotics, American Association for the Advancement of Science, v. 3, n. 21, eaat5954, 2018.

BERS, Marina Umaschi; FLANNERY, Louise; KAZAKOFF, Elizabeth R; SULLIVAN, Amanda. Computational thinking and tinkering: Exploration of an early childhood robotics curriculum. Computers & education, Elsevier, v. 72, p. 145–157, 2014.

BUJANG, Mohamad Adam; OMAR, Evi Diana; BAHARUM, Nur Akmal. A review on sample size determination for Cronbach’s alpha test: a simple guide for researchers. The Malaysian journal of medical sciences: MJMS, v. 25, n. 6, p. 85, 2018.

EFTHYMIOU, A.; BAX, I.; KESSLER, T. ChildBot: Multi-Robot Perception and Interaction with Children. arXiv preprint arXiv:2002.05419, 2020.

GÓMEZ RODRÍGUEZ, L. Robótica educativa utilizando el mBot en estudiantes de educación básica. Revista Iberoamericana de Tecnología en Educación y Educación en Tecnología, SciELO México, v. 29, p. 67–78, 2022.

HADFIELD, K.; TAYLOR, H.; BELPAEME, T. A Deep Learning Approach for Multi-View Engagement Estimation of Children in a Child-Robot Joint Attention task. arXiv preprint arXiv:1803.03163, 2018.

KRIZHEVSKY, Alex. Learning multiple layers of features from tiny images. [S. l.: s. n.], 2009.

KULKARNI, Uday; MEENA, SM; GURLAHOSUR, Sunil V; BHOGAR, Gopal. Quantization friendly mobilenet (qf-mobilenet) architecture for vision based applications on embedded platforms. Neural Networks, Elsevier, v. 136, p. 28–39, 2021.

MOHD, Bassam J; AHMAD YOUSEF, Khalil M; ALMAJALI, Anas; HAYAJNEH, Thaier. Quantization-Based Optimization Algorithm for Hardware Implementation of Convolution Neural Networks. Electronics, MDPI, v. 13, n. 9, p. 1727, 2024.

NOVAC, Pierre-Emmanuel; BOUKLI HACENE, Ghouthi; PEGATOQUET, Alain; MIRAMOND, Benoit; GRIPON, Vincent. Quantization and deployment of deep neural networks on microcontrollers. Sensors, MDPI, v. 21, n. 9, p. 2984, 2021.

RAMÍREZ SÁNCHEZ, C.; LANDÍN JUÁREZ, M. Modelo de Robótica Educativa con el Robot Darwin Mini para Desarrollar Competencias en Estudiantes de Licenciatura. Revista Electrónica de Investigación Educativa, SciELO México, v. 19, n. 2, p. 45–60, 2017.

SONG, Zhuoran; FU, Bangqi; WU, Feiyang; JIANG, Zhaoming; JIANG, Li; JING, Naifeng; LIANG, Xiaoyao. DRQ: Dynamic region-based quantization for deep neural network acceleration. In: IEEE. 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA). [S. l.: s. n.], 2020. p. 1010–1021.

UNESCO. Informe de Seguimiento de la Educación en el Mundo 2024. [S. l.: s. n.], 2024. Accessed: 2025-03-27. Available from: https://unesdoc.unesco.org/ark:/48223/pf0000391983.

VINUESA, Ricardo; AZIZPOUR, Hossein; LEITE, Iolanda; BALAAM, Madeline; DIGNUM, Virginia; DOMISCH, Sami; FELLÄNDER, Anna; LANGHANS, Simone D; TEGMARK, Max; FUSO NERINI, Francesco. The role of artificial intelligence in achieving the Sustainable Development Goals. Nature Communications, Nature Publishing Group, v. 11, n. 1, p. 1–10, 2020. DOI: 10.1038/s41467-019-14108-y.

VYGOTSKY, Lev S. Mind in society: The development of higher psychological processes. [S. l.]: Harvard university press, 1978. v. 86.

WARDEN, Pete; SITUNAYAKE, Daniel. Tinyml: Machine learning with tensorflow lite on arduino and ultra-low-power microcontrollers. [S. l.]: O’Reilly Media, 2019.

XIAO, Penghao; ZHANG, Chunjie; GUO, Qian; XIAO, Xiayang; WANG, Haipeng. Neural networks integer computation: Quantizing convolutional neural networks of inference and training for object detection in embedded systems. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, IEEE, 2024.