To truly understand the visual world our models should be able not only to recognize images but also generate them. To this end, there has been exciting recent progress on generating images from natural language descriptions... (read more)
PDFMETHOD | TYPE | |
---|---|---|
![]() |
Convolutions |