NAS-DIP: Learning Deep Image Prior with Neural Architecture Search
Back
Results on single image super-resolution
Dataset:
Set5
Set14
Resolution:
4x
8x
Frame:
01
/
5
: Toy
Method:
GT
SelfExSR
LapSRN
VDSR
DIP
Ours
The slider shows images between the input (left) and the results (right).
To compare between different methoods, please slide the bar to the left and click different tabs.
Set5 dataset link
Input sequence