摘要
We propose an unfolding network GapUNet for spatial-temporal compressive imaging. Simulation and optical experiments demonstrate the network performance using compression ratios of 128: 1 and 16: 1. The mean PSNR of the reconstructed objects is higher than 29dB.
源语言 | 英语 |
---|---|
期刊 | Optics InfoBase Conference Papers |
出版状态 | 已出版 - 2022 |
活动 | Computational Optical Sensing and Imaging, COSI 2022 - Vancouver, 加拿大 期限: 11 7月 2022 → 15 7月 2022 |