I want to know whether changing the data into. H5 file will affect the effect of network training? I used my own data set for training, and I didn't convert the image into. H5 file, but after preprocessing the data, I input it into the network. The network uses the network you proposed, but why does the output result have mosaic effect? The effect is very bad?
I want to know whether changing the data into. H5 file will affect the effect of network training? I used my own data set for training, and I didn't convert the image into. H5 file, but after preprocessing the data, I input it into the network. The network uses the network you proposed, but why does the output result have mosaic effect? The effect is very bad?