Title	:	Handling Motion Blur in Multi-Frame Super-Resolution
Authors	: 	Ziyang Ma, Renjie Liao, Xin Tao, Li Xu, Jiaya Jia, Enhua Wu

Supplementary Material
=============================================================

IMG_****\			- low-res images directly cropped from sequences captured using an iPhone.

IMG_****\results_**\ 		- our results as well as other state-of-the-arts.

IMG_****_addSyntheticNoise\ 	- low-res images with synthetically random noise; our algorithm is robust to noise.

woman\				- images from surveillance video.

-------------------------------------------------------------
In the "results\" folder, files are named using the following convention:

img_xxxxx_bicubic.png   - the result of a bicubically upsampled input frame that we want to super-resolve
our final result.png    - the result generated using our algorithm presented in this paper
liu et al.png           - the result without handling motion blur (Liu & Sun [2])
shan et al.png          - the result of video upsampling (shan et al [3])
xu et al.png            - the result of single image deblurring (xu et al [5])
zhu et al.png           - the result of a recent multi-image deblurring (zhu et al [7]) + multi-frame SR (Liu & Sun [2])
zhang et al.png         - the result of a recent multi-shot imaging method (zhang et al [6])
-------------------------------------------------------------
NOTE: Not all results given for some sequences, because either the authors did not run the data for us, or they have appeared in the supplementary PDF. 

Please visit our website 'http://www.cse.cuhk.edu.hk/leojia/projects/mfsr' for more data and results.
