The field of image and video enhancement is rapidly evolving, with a focus on developing innovative methods to improve the quality and resolution of visual data. Recent research has explored the application of deep learning techniques, such as convolutional neural networks (CNNs) and transformers, to achieve state-of-the-art results in image super-resolution, video deblurring, and other related tasks. Noteworthy papers in this area include Native-Resolution Image Synthesis, which introduces a novel generative modeling paradigm for synthesizing images at arbitrary resolutions and aspect ratios, and DualX-VSR, which proposes a dual axial spatial-temporal transformer for real-world video super-resolution without motion compensation. These advancements have significant potential for applications in various fields, including computer vision, robotics, and healthcare.
Advancements in Image and Video Enhancement
Sources
Application of convolutional neural networks in image super-resolution
Native-Resolution Image Synthesis
WIFE-Fusion:Wavelet-aware Intra-inter Frequency Enhancement for Multi-model Image Fusion
SAAT: Synergistic Alternating Aggregation Transformer for Image Super-Resolution
Joint Video Enhancement with Deblurring, Super-Resolution, and Frame Interpolation Network
Video Deblurring with Deconvolution and Aggregation Networks
On the Synthetic Channels in Polar Codes over Binary-Input Discrete Memoryless Channels
HMAR: Efficient Hierarchical Masked Auto-Regressive Image Generation
EECD-Net: Energy-Efficient Crack Detection with Spiking Neural Networks and Gated Attention
Enhancing Frequency for Single Image Super-Resolution with Learnable Separable Kernels
MARS: Radio Map Super-resolution and Reconstruction Method under Sparse Channel Measurements
DualX-VSR: Dual Axial Spatial$\times$Temporal Transformer for Real-World Video Super-Resolution without Motion Compensation
Geological Field Restoration through the Lens of Image Inpainting
Multi-scale Image Super Resolution with a Single Auto-Regressive Model
CSI2Vec: Towards a Universal CSI Feature Representation for Positioning and Channel Charting
AliTok: Towards Sequence Modeling Alignment between Tokenizer and Autoregressive Model