Inpainting
PulseAugur coverage of Inpainting — every cluster mentioning Inpainting across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New EBM-AE Framework Enhances Generative Modeling and Image Inpainting
Researchers have developed a novel cooperative framework that merges Energy-Based Models (EBMs) with Autoencoders (AEs) to enhance generative modeling. This EBM-AE framework employs an iterative process where an EBM ref…
-
SDXL Inpainting with Multiple LoRAs: User Seeks Real-Time Functionality
A user on Reddit is inquiring about the possibility of using multiple LoRAs simultaneously for inpainting with SDXL models. They recall seeing a demonstration of near real-time inpainting with character and style LoRAs …
-
New method integrates aleatoric and epistemic uncertainty in deep learning
Researchers have developed a new method for uncertainty quantification (UQ) in deep learning models, particularly for high-dimensional output spaces. This approach jointly models aleatoric uncertainty, which accounts fo…
-
New methods enhance AI's ability to solve imaging inverse problems
Researchers have developed new methods, Spectrum-Adaptive Scheduling (SAS) and Measurement-Prioritized Attention (MPA), to improve the performance of flow-based generative models in solving imaging inverse problems. The…
-
New frameworks enhance mask transformers and adapt State Space Models for missing data
Researchers have developed iFAN, a training framework designed to enhance mask transformers by aligning query ranking with mask quality and improving intermediate prediction distillation. This method addresses mismatche…
-
New GAFIC method enhances image cropping with global attention
Researchers have developed a new image cropping method called Global Attention-Fused Image Cropping (GAFIC) that aims to improve aesthetic composition by considering global relationships between image components, rather…
-
Fast Equivariant Imaging accelerates unsupervised deep learning training
Researchers have introduced Fast Equivariant Imaging (FEI), a new unsupervised learning framework designed to accelerate the training of deep imaging networks. FEI reformulates the Equivariant Imaging objective using an…
-
New methods enhance visual autoregressive models for image generation
Researchers have developed new methods to improve the efficiency and performance of visual autoregressive models. One approach, Shift-and-Sum Quantization, addresses reconstruction errors in attention-value products and…
-
Research Unifies Data-Driven Priors for Bayesian Inverse Problems
A new research paper proposes a unified framework for integrating various data-driven priors into Bayesian inverse problems. The study demonstrates how diverse priors, including regularization-by-denoising, normalizing …
-
Image inpainting research highlights reward model biases
Researchers have re-examined preference alignment for image inpainting, utilizing the Direct Preference Optimization framework with publicly available reward models. Their study revealed that while most reward models of…