VFHQ: A High-Quality Dataset and Benchmark for Video Face Super-Resolution.
Liangbin XieXintao WangHonglun ZhangChao DongYing ShanPublished in: CVPR Workshops (2022)
Keyphrases
- super resolution
- high quality
- video super resolution
- high resolution
- low resolution
- single frame
- high resolution video
- higher resolution
- image reconstruction
- low resolution images
- face images
- image restoration
- video data
- image super resolution
- depth map
- highest resolution
- combining information from multiple
- super resolved
- motion blur
- motion estimation
- video sequences
- resolution enhancement
- multi frame
- human faces
- super resolution reconstruction
- image resolution
- space time
- face hallucination
- multimedia
- high resolution face images
- light field
- higher quality
- facial features
- image quality
- high resolution images
- temporal resolution
- high spatial resolution
- field of view
- image patches
- video frames
- image classification
- markov random field
- spatio temporal
- training set