Showing 2023Show all
2 papers · 1 filter
cs.CV2023
Shadows Don't Lie and Lines Can't Bend! Generative Models don't know Projective Geometry...for now
Ayush Sarkar, Hanlin Mai, Amitabh Mahapatra +3
Generative models can produce impressively realistic images. This paper demonstrates that generated images have geometric features different from those of real images. We build a s…
cs.CV2023
Improving Equivariance in State-of-the-Art Supervised Depth and Normal Predictors
Yuanyi Zhong, Anand Bhattad, Yu-Xiong Wang +1
Dense depth and surface normal predictors should possess the equivariant property to cropping-and-resizing -- cropping the input image should result in cropping the same output ima…