Deep Learning

Contextualizing Bandera

The goal of this project is twofold: First, it is to be con­sid­ered a con­tri­bu­tion to the emerg­ing field of “dis­tant view­ing”, which uses quan­ti­ta­tive meth­ods to assess a cor­pus con­sist­ing of a large num­ber of visu­al media. Cur­rent­ly, deep learn­ing meth­ods play a minor role in dis­tant view­ing, as most of the projects use pre-trained net­works. This is under­stand­able, as train­ing is not triv­ial. How­ev­er, using pre-trained net­works sig­nif­i­cant­ly reduces the amount of pos­si­ble research ques­tions. More­over, a bet­ter under­stand­ing of the train­ing process allows us to con­tribute to the field of “crit­i­cal machine learn­ing” as well; specif­i­cal­ly we try to point out some of the ben­e­fits and pit­falls of train­ing an arti­fi­cial neur­al net­work for a human­i­ties research project.

Sec­ond, this project extends the method­ol­o­gy of study­ing cul­tur­al mem­o­ry. By means of train­ing a resid­ual con­vo­lu­tion­al neur­al net­work in Facebook’s Detectron2 frame­work to rec­og­nize a num­ber of national(istic) sym­bols from East­ern Europe, we show how Ukrain­ian nation­al­ist Stepan Ban­dera (1909–1959) is instru­men­tal­ized in the online dis­course about the recent Ukrain­ian cri­sis. From 2013 onwards we have col­lect­ed a total of 800 YouTube video clips about Ban­dera. Our cus­tom-trained net­work is then used to iden­ti­fy the respec­tive national(istic) con­text of the videos in question.