Sora-generated videos associate stereotyped occupations, behaviors, and appearances with specific genders, based on frequency counts from 120 clips.
The findings reveal significant biases that align with traditional gender stereotypes, with females being predominantly associated with traits such as attractiveness and frailty
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Gender Bias in Text-to-Video Generation Models: A case study of Sora
Sora-generated videos associate stereotyped occupations, behaviors, and appearances with specific genders, based on frequency counts from 120 clips.