This repo reproduces a prototype of the classical algorithm for Single Image Super resolution (SISR) with convolutional neural network based on MXNET.

• MXNET 0.11.0
• MKL/Numpy/h5py/scipy

• CPU version of MXNET
• Win 10 64-bit & 8G

• The uploaded checkpoint may not provide high performance since only 1,000 iterations were done.

• Dong C, Loy C C, He K, et al. Image super-resolution using deep convolutional networks[J]. IEEE T. PAMI, 2016, 38(2): 295-307.