Rethinking image super resolution from long-tailed distribution learning perspective
Existing studies have empirically observed that the resolution of the low-frequency region is
easier to enhance than that of the high-frequency one. Although plentiful works have been …
easier to enhance than that of the high-frequency one. Although plentiful works have been …
Combination of super-resolution reconstruction and SGA-Net for marsh vegetation map** using multi-resolution multispectral and hyperspectral images
Vegetation is crucial for wetland ecosystems. Human activities and climate changes are
increasingly threatening wetland ecosystems. Combining satellite images and deep …
increasingly threatening wetland ecosystems. Combining satellite images and deep …
Learning re-sampling methods with parameter attribution for image super-resolution
Single image super-resolution (SISR) has made a significant breakthrough benefiting from
the prevalent rise of deep neural networks and large-scale training samples. The …
the prevalent rise of deep neural networks and large-scale training samples. The …
Data augmentation for multi-image super-resolution
Super-resolution reconstruction consists in generating a high-resolution image from a single
low-resolution image or multiple images presenting the same area of interest. Existing state …
low-resolution image or multiple images presenting the same area of interest. Existing state …
RepCaM: Re-parameterization Content-aware Modulation for Neural Video Delivery
Recently, content-aware methods have been utilized to reduce the bandwidth and improve
the quality of Internet video delivery. Existing methods train corresponding content-aware …
the quality of Internet video delivery. Existing methods train corresponding content-aware …
DDA: A dynamic difficulty-aware data augmenter for image super-resolution
Deep neural networks (DNNs) have been recently widely used in image super-resolution
(SR) and have achieved remarkable performance. However, most existing methods focus on …
(SR) and have achieved remarkable performance. However, most existing methods focus on …
Rethinking Imbalance in Image Super-Resolution for Efficient Inference
Existing super-resolution (SR) methods optimize all model weights equally using $\mathcal
{L} _1 $ or $\mathcal {L} _2 $ losses by uniformly sampling image patches without …
{L} _1 $ or $\mathcal {L} _2 $ losses by uniformly sampling image patches without …