These papers focus on all aspects of the theory and practice of imaging systems, exploring work relating to the algorithmic development, design, implementation and application of image and video technology.
Explore the following trending articles in Frontiers in Signal Processing.
This article presents a taxonomy of immersive communication systems (ICS), categorizing them based on perceptual features. It also evaluates existing user experience (QoE) assessment methods and proposes guidelines for robust evaluation.
This comprehensive survey explores deep learning approaches to point cloud compression, crucial for storage and transmission as demand rises. It analyzes challenges, current methods, and their interplay with traditional compression techniques.
AL-Net, a lightweight network, balances segmentation accuracy and inference speed for medical images. It employs RepVGG-A1 for feature extraction, Lite R-ASPP for context, and an asymmetric decoder for performance enhancement.
LVAC, an end-to-end compression framework, targets volumetric functions, especially point cloud attributes. By tiling space into blocks and using shifts of a CBN, it achieves superior compression, promising broader applications beyond point clouds.
This study proposes a method to detect GAN-generated face images using inter-eye symmetries and inconsistencies. Employing a Siamese Neural Network, it achieves robust detection across various GAN models and conditions.
The V3C standard enables efficient coding, storage, and transport of immersive content like volumetric video. This article outlines its concepts, storage, streaming, and integration with existing infrastructures.
The article introduces a modular RGB-D capture system for photorealistic 3D user representations in XR communication. Technical evaluations and user studies validate its feasibility and effectiveness in immersive applications.
This paper introduces MRET, a Transformer-based framework for NR-VQA of UGC videos. It preserves high-resolution details and effectively captures global composition and local details for robust performance across resolutions.
Perceptual Video Quality Assessment (VQA) is crucial in Video Engineering, evolving significantly with video compression. VQA algorithms have shifted from hand-crafted to deep learning methods.
Researchers propose novel methods for blind restoration of low-quality faces. They introduce Degradation Learning (DL) via style transfer and a hybrid architecture called Skip-Transformer (ST).
Read other impactful articles published in Frontiers in Signal Processing.