Performance Improvement of Deblurring Using Learning Methods
Tatsuya Mizuno, Tomio Goto · 2022 IEEE 11th Global Conference on Consumer Electronics (GCCE) · 2022
Machine learning recovery performance tends to be poor for data not included in the data set. Model performance is improved by using various kind of blur image for learning image in deblurring. We obtained on average 1.038 dB higher results in PSNR, 0.012 dB higher results in SSIM. We could make more versatile method.