DIGITAL LIBRARY
DIFFUSION-MODELS AS INPAINTING INQUIRIES FOR EXPRESSIVE-ART BASED FORMATIVE-ASSESSMENT
Ilmenau University of Technology (GERMANY)
About this paper:
Appears in: ICERI2024 Proceedings
Publication year: 2024
Pages: 4983-4992
ISBN: 978-84-09-63010-3
ISSN: 2340-1095
doi: 10.21125/iceri.2024.1228
Conference name: 17th annual International Conference of Education, Research and Innovation
Dates: 11-13 November, 2024
Location: Seville, Spain
Abstract:
This work-in-progress paper ‎presents our expressive-art based Formative-Assessment (FA) practice ‎as inquiry in intercultural settings for liberation STEM education. The settings focus on target ‎groups of repressed adult-migrants/students, migrated to Germany, due to the enduring ‎oppressive structures and the current wars in Ukraine and Syria. The target groups are interested ‎in experimenting and programming Diffusion-Model (DM) inpainting algorithms to restore their ‎damaged cultural or/and personal images due to the wars.‎

The paper shows how the practice is conceptualized to enable two DM prototypes:
1) A prototype for Ukrainian target group: It is conceptualized to serve a target group of Ukraine students and help them experiment DM inpainting algorithms to face and overcome ‎the repressive barriers in Ukrainian cultural context.
2) A prototype for Syrian target group: It is conceptualized to serve a target group of Syrian students and ‎help them experiment DM inpainting algorithms to face and overcome the repressive barriers in Syrian cultural context.

For enabling our expressive-art ‎based FA workshops to approach the cultural aspect of repression in context of each target ‎group, the paper shows how the workshops’ two DM prototypes are grounded based on the ‎theoretical concepts “Generative words”, “Generative themes”, “Critical consciousness” and ‎‎“Reading the world precedes reading the word” suggested by the critical educator Paul Freire [1,2,3]. ‎

The paper also shows the differences between the DM inpainting algorithms used to initialize and build the two DB prototypes and the deep-‎learning and machine learning algorithms used in our previous research works [4,5,6,7,8]. ‎

To test the two DM prototypes, test cases are suggested as next steps of this work-in-‎progress paper. Finally, for reflection and feedback exchange with the community we ‎summarize the limits of the two DM prototypes and our next steps to improve them. ‎

References:
‎[1] Frerie; Paulo, Pedogogy of Freedom; Ethics, Democracy, and Civic Courage; Rowman & ‎Littlefield Publishers; ‎First Paperback Printing (2001)‎
‎[2] Freire, Paulo. 1985, The Politics of Education: Culture, Power and Liberation, Copyright© ‎‎1985 by Bergin & ‎Garvey Publishers, Inc ‎
‎[3] Freire, Paulo, Pedagogy of the Oppressed. Harmondsworth: Penguin (1972).‎
‎[4]‎ W. Mahfouz, H. -D. Wuttke and S. Werner, "Formative-Assessment Freirean-Dialogue API ‎for Data-Analytics Trainee-Teachers,"2023 IEEE Frontiers in Education Conference (FIE), ‎College Station, TX, USA, 2023, pp. 1-5, doi: 10.1109/FIE58773.2023.10342969.‎
‎[5]‎ W. Mahfouz and H. -D. Wuttke, "Deep-learning API-feature Specification Tool for ‎Formative Assessment in Workshops," 2022 IEEE Frontiers in Education Conference (FIE), ‎Uppsala, Sweden, 2022, pp. 1-9, doi: 10.1109/FIE56618.2022.9962474. ‎
‎[6]‎ W. Mahfouz, H.D. Wuttke (2022) INTEGRATING FLOW THEORY IN FREIREAN ‎PEDAGOGY BASED WORKSHOPS ‎FOR FORMATIVE ASSESSMENT, ICERI2022 ‎Proceedings, pp. 6299-6309. ‎
‎[7]‎ W. Mahfouz and H. -D. Wuttke, "Big Data Analytics APIs Architecture for Formative ‎Assessors," 2021 IEEE Frontiers in Education Conference (FIE), Lincoln, NE, USA, 2021, pp. ‎‎1-9, doi: 10.1109/FIE49875.2021.9637431.‎
‎[8]‎ W. Mahfouz and H. -D. Wuttke, "Automatic Classifiers for Formative Assessment," 2019 ‎IEEE Frontiers in Education Conference (FIE), Covington, KY, USA, 2019, pp. 1-9, doi: ‎‎10.1109/FIE43999.2019.9028713. ‎
Keywords:
Formative Assessment, Liberation STEM ‎‎education, Expressive Art.