no code implementations • 29 Nov 2022 • HyeonCheol Moon, Jinwoo Jeong, Sungjei Kim
More recently, KD for SISR utilizes Feature Distillation (FD) to minimize the Euclidean distance loss of feature maps between teacher and student networks, but it does not sufficiently consider how to effectively and meaningfully deliver knowledge from teacher to improve the student performance at given network capacity constraints.
Ranked #10 on Image Super-Resolution on Set5 - 3x upscaling