Latent diffusion model
PulseAugur coverage of Latent diffusion model — every cluster mentioning Latent diffusion model across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
LungMend corrects TB CT lung masks using healthy-only diffusion model
Researchers have developed LungMend, a novel post-processing model designed to correct lung segmentation masks in computed tomography (CT) scans affected by severe tuberculosis. Unlike existing models that struggle with…
-
AI synthesizes controllable glioblastoma MRI scans using latent diffusion models
Researchers have developed a novel two-stage latent diffusion model (LDM) framework designed to synthesize controllable 2D glioblastoma MRI scans. The first stage generates tumor segmentation maps based on specified loc…
-
AI synthesizes 3D pathology data from 2D images to improve cancer classification
Researchers have developed PathoS2VD, a generative framework designed to create pseudo Z-stack representations from single 2D histopathology images. This approach aims to capture morphological information along the z-ax…
-
LatentDiff framework enhances knowledge graph completion via latent diffusion
Researchers have introduced LatentDiff, a novel framework for triple set prediction in knowledge graphs. This approach reformulates knowledge graph completion as a latent space graph generation problem. LatentDiff utili…
-
FlowMesh framework advances physical field modeling on unstructured meshes
Researchers have developed FlowMesh, a novel generative framework designed for stable distributional modeling of physical fields on unstructured meshes. This approach combines permutation-invariant mesh encoding with de…
-
New diffusion models enhance medical imaging and general image generation
Researchers have developed new diffusion model techniques for medical imaging and general image generation. One approach, MC-LDM, uses multimodal data like MRI and PET scans to forecast Alzheimer's disease progression, …
-
New framework curates synthetic data for scarce few-shot learning
Researchers have introduced a new framework for Generalizable Few-Shot Class-Incremental Learning (G-FSCIL), a scenario where both base and incremental data are scarce. The proposed method addresses challenges like weak…
-
StyleGAN2 enhances geological data assimilation via intermediate latent space
Researchers have developed a novel parameterization method for ensemble-based data assimilation using the intermediate latent space of StyleGAN2. This approach addresses limitations of traditional methods that struggle …
-
AI generates culturally faithful Ulos motifs using multimodal Stable Diffusion XL
Researchers have developed a multimodal generative framework to aid the traditional Batak Ulos weaving industry by enabling controllable and culturally faithful motif generation. The system fine-tunes Stable Diffusion X…
-
User offers local testing for older AI image generation models
A Reddit user has successfully set up and is offering to test prompts on a variety of older AI image generation models locally. These models include First Order Motion Model, DeepDaze, Big Sleep, VQGAN+CLIP, OG DALL-E M…
-
GRADE system uses generative AI to improve radar depth estimation in poor visibility
Researchers have developed GRADE, a novel system for estimating high-fidelity metric depth from single-frame radar data, particularly under challenging visual conditions like smoke, fog, and darkness. GRADE leverages a …
-
New EEGDM framework uses latent diffusion models for signal representation
Researchers have introduced EEGDM, a novel self-supervised framework designed to learn representations from electroencephalogram (EEG) data. Unlike previous methods that focus on reconstructing masked signal segments, E…
-
New AI models ReFlowSET and C-DiffSET advance SAR-to-EO image translation
Researchers have developed two new frameworks, ReFlowSET and C-DiffSET, for translating synthetic aperture radar (SAR) images into electro-optical (EO) imagery. ReFlowSET focuses on selecting an optimal latent codec and…
-
New framework enhances fairness in AI face generation
Researchers have developed a new framework called Semantic Boundary Predictor (SBP) to improve fairness in synthetic face generation using latent diffusion models. SBP intervenes once during the reverse diffusion proces…
-
New method uses metadata to guide synthetic cardiac MRI generation
Researchers have developed a new method for generating synthetic cardiac magnetic resonance imaging (CMR) using a pre-trained latent diffusion model. This approach conditions the model on structured clinical metadata an…
-
On-device diffusion app uses ambient light instead of text prompts
Researchers have developed an Android application that performs latent diffusion image generation entirely on-device, bypassing cloud services for enhanced auditability. This application utilizes the device's ambient-li…
-
DiffGI introduces differentiable geometry images for high-fidelity 3D generation
Researchers have introduced DiffGI, a novel framework for 3D generation that utilizes differentiable geometry images. This approach replaces traditional volumetric representations with continuous 2D Truncated Signed Dis…
-
DiffGI framework enables high-fidelity thin-shell 3D generation
Researchers have introduced DiffGI, a novel framework for generating high-fidelity 3D models, particularly effective for thin-shell structures like garments. Unlike previous methods that used discrete occupancy maps, Di…
-
Quantum circuits show promise and challenges in AI generative models
Researchers are exploring the integration of quantum circuits into AI models, particularly for generative tasks like image synthesis and quantum circuit optimization. One study on quantum circuit synthesis found that wh…
-
CONFLUX model generates realistic 3D chest CT scans with enhanced clinical control
Researchers have developed CONFLUX, a novel latent diffusion model designed for synthesizing 3D chest CT scans with specific clinical attributes. The model utilizes a 3D variational autoencoder for compression and a rec…