peak signal-to-noise ratio
PulseAugur coverage of peak signal-to-noise ratio — every cluster mentioning peak signal-to-noise ratio across labs, papers, and developer communities, ranked by signal.
13 day(s) with sentiment data
-
JAX3D enables hierarchical NeRF for advanced 3D rendering and reconstruction
Researchers have developed a method for creating hierarchical Neural Radiance Fields (NeRFs) using JAX and the jax3d library. This approach enables volumetric rendering, novel-view synthesis, and 3D reconstruction. The …
-
New LaST-SR method enhances image super-resolution with complex-frequency decomposition
Researchers have introduced LaST-SR, a novel single image super-resolution method that utilizes a complex-frequency decomposition approach. This technique combines a global Fourier branch for broad image context and a l…
-
New decomposition method enhances image restoration with cone constraints
Researchers have developed a new method for image restoration using a cone-constrained bilinear decomposition of the total scaled-gradient variation (TSGV) regularizer. This approach addresses the computational challeng…
-
Medical foundation models enhance brain MRI contrast dose simulation
Researchers have developed a new method for simulating brain MRI contrast doses by utilizing features from medical foundation models as a perceptual loss. This approach aims to improve the accuracy of image synthesis co…
-
Study finds deep learning MRI reconstruction models lack safety evaluation
A recent study published on arXiv evaluated the safety of deep learning models used for brain MRI reconstruction. The research found that current evaluation methods, which often rely on metrics like PSNR and SSIM, are i…
-
PixelIR framework decouples image fidelity and perception for super-resolution
Researchers have introduced PixelIR, a novel framework for image super-resolution that decouples fidelity and perceptual quality. Unlike previous methods that optimize both objectives simultaneously, PixelIR first gener…
-
New VTON evaluation framework DAT outperforms Gemini, Qwen, and GPT-5.5
Researchers have developed a new framework called DAT to evaluate virtual try-on (VTON) models more effectively. Existing metrics like FID and SSIM struggle to capture garment fidelity, so DAT breaks down consistency in…
-
Research: Projection choice impacts learned video compression for 360-degree content
A new research paper explores how different projection formats impact the efficiency of end-to-end learned video compression for 360-degree content. The study found that equirectangular and padded equirectangular projec…
-
New ABCD framework enables large radiance field training on limited VRAM
Researchers have introduced ABCD (Alpha-Composited Block Coordinate Descent), a novel out-of-core training framework designed for large radiance fields, specifically demonstrated with 3D Gaussian Splatting. This method …
-
New framework tackles continual learning for image restoration
Researchers have developed a novel framework called Restoring Without Forgetting (RwF) to address the challenge of continual learning in image restoration tasks. This framework enables models to adapt to new image degra…
-
New autoencoder method quantifies image differences using latent representations
Researchers have developed a new method for quantifying image differences using autoencoder-based latent representations. This approach leverages deep neural networks to capture high-level semantic information, offering…
-
New AI model enhances low-light drone imagery for bridge damage detection
Researchers have developed DaL-MoE, a new image restoration technique designed to improve bridge damage detection in low-light conditions using unmanned aerial vehicles (UAVs). This method employs an ISP-aware synthesis…
-
New GS-Net module boosts autonomous vehicle data reuse with faster 3DGS
Researchers have developed GS-Net, a novel module designed to enhance data reuse across different autonomous vehicles. This plug-and-play system aggregates geometric context from sparse Structure-from-Motion point cloud…
-
AutoLumNet framework offers state-of-the-art single-shot image exposure correction
Researchers have introduced AutoLumNet, a novel framework designed for single-shot exposure correction in images. This system decomposes the correction process into a global monotone tone curve and a local residual, ens…
-
Quantum-inspired TT-Net advances image denoising with tensor networks
Researchers have introduced TT-Net, a novel approach for image denoising that leverages quantum-inspired tensor network methods. Unlike existing methods that use singular value decomposition (SVD) on individual channels…
-
SFMformer achieves SOTA image super-resolution with novel Transformer design
Researchers have developed SFMformer, a new lightweight Transformer model for image super-resolution that achieves state-of-the-art results. The model utilizes a novel spatial-frequency modulation approach, combining sp…
-
New SAR despeckling method achieves top performance in benchmarks
Researchers have developed a novel method for synthetic aperture radar (SAR) despeckling, a process that removes noise from SAR images without obscuring important scattering structures. The new technique revisits a nonl…
-
New 4D-SG Method Enhances Sparse-View Spectral CT Reconstruction
Researchers have developed a new method called Shared-Structure 4D Spectral Gaussian Representation (4D-SG) for reconstructing energy-resolved attenuation volumes from limited computed tomography (CT) projection views. …
-
New Gaussian Volume Encoding Method Achieves High Compression
Researchers have developed a novel method for encoding scalar volumes using anisotropic Gaussian primitives under a fixed budget. This structure-aware allocation technique extracts positional, orientational, and shape i…
-
New framework anchors black-box denoiser output for improved fidelity
Researchers have developed a new framework called Fidelity-Constrained Anchoring designed to improve the output of black-box denoisers. This method blends the denoised image with the original input, applying a blending …