LSHR-Net: A hardware-friendly solution for high-resolution computational imaging using a mixed-weights neural network.
Fangliang BaiJinchao LiuXiaojuan LiuMargarita OsadchyChao WangStuart J. GibsonPublished in: Neurocomputing (2020)
Keyphrases
- high resolution
- neural network
- image processing
- low resolution
- super resolution
- artificial neural networks
- hardware and software
- computational power
- data acquisition
- synthetic aperture sonar
- dedicated hardware
- video rate
- satellite images
- field of view
- back propagation
- high quality
- neural network model
- medical imaging
- network architecture
- high frequency
- numerical methods
- high resolution images
- optimal solution
- magnetic resonance images
- hidden layer
- weighted sum
- higher resolution
- synthetic aperture
- hidden neurons