mdctGAN: Taming transformer-based GAN for speech super-resolution with Modified DCT spectra.
Chenhao ShuaiChaohua ShiLu GanHongqing LiuPublished in: CoRR (2023)
Keyphrases
- super resolution
- low resolution
- high resolution
- image restoration
- image reconstruction
- motion estimation
- high quality
- low resolution images
- image compression
- resolution enhancement
- high resolution images
- image super resolution
- multi frame
- super resolution reconstruction
- light field
- motion blur
- super resolved
- speech signal
- discrete cosine transform
- image resolution
- higher resolution
- super resolution algorithm
- hyperspectral images
- spatial resolution
- structuring elements
- single image super resolution
- single frame
- transform domain
- spatial domain
- motion vectors
- depth map
- dct domain
- dct coefficients
- compressed domain
- computer vision
- multispectral
- principal component analysis
- computational complexity
- high resolution video