Cycle-consistent generative adversarial networks (CycleGAN) have shown their promising performance for speech enhancement (SE), while one intractable shortcoming of these CycleGAN-based SE systems is that the noise components propagate throughout cycle and cannot be completely eliminated. Additionally, conventional only estimate spectral magnitude, phase unaltered. Motivated by multi-stage lear...

AbstractLow-light image enhancement is a challenging low-level computer vision task because after enhancing the brightness of image, amplified noise, color distortion, detail loss, blurred edges, shadow blocks and halo artifacts occur. In this paper, we propose Two-Stage Network with Channel Attention (denoted as TSN-CA) to enhance low-light restore enhanced images from various kinds degradatio...

data envelopment analysis is a nonparametric technique checking efficiency of dmus using math programming. in conventional dea, it has been assumed that the status of each measure is clearly known as either input or output. kao and hwang (2008) developed a data envelopment analysis (dea) approach for measuring efficiency of decision processes which can be divided into two stages. the first stag...

data envelopment analysis (dea) is a method for measuring the efficiency of peer decision making units (dmus) which uses a set of inputs to produce a set of outputs. in some cases, dmus have a two-stage structure, in which the first stage utilizes inputs to produce outputs used as the inputs of the second stage to produce final outputs. one important issue in two-stage dea is the sensitivity of...