Channel Attention-Based Approach with Autoencoder Network for Human Action Recognition in Low-Resolution Frames.
Elaheh DastbaravardehSomayeh AskarpourMaryam Saberi AnariKhosro RezaeePublished in: Int. J. Intell. Syst. (2024)
Keyphrases
- low resolution
- high resolution
- super resolution
- single frame
- low resolution images
- high quality
- high resolution images
- face images
- image super resolution
- resolution enhancement
- lower resolution
- higher resolution
- low resolution face images
- super resolved
- dense correspondences
- image processing
- multi frame
- low quality images
- image patches
- low frame rate
- high resolution video
- peripheral vision
- pixel intensities
- super resolution algorithm
- super resolution reconstruction
- frame rate
- high speed
- motion estimation
- spatio temporal
- three dimensional