How image labeling is done?
#1
by
Rhinobot
- opened
I want to contribute and work on this dataset. How sketch conversion and caption are generated from dresscode dataset?
Hi the image labelling was done partially by manual labelling and partly by a fine tuned model using CLIP by the researchers at University of Milan, for more information read this paper https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136680337.pdf