It's required to find the best image given with a text description. However, the resolution is very low, i.e., 50 x 50 pixels.
In this case, can pre-trained ResNet50 be used? or any recommendations on a better architecture? Thanks!
Asked
Active
Viewed 231 times
