3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation
PulseAugur coverage of 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation — every cluster mentioning 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New PACK3D attention operator improves 3D data handling in AI models
Researchers have introduced PACK3D, a novel attention operator designed to address the challenges of packing 3D volumetric data into token sequences for AI models. Traditional methods struggle because sequence order can…
-
New AI protocol struggles to predict future MS lesions beyond population averages
Researchers have developed FORESEE-MS, a new evaluation protocol aimed at predicting future multiple sclerosis (MS) lesions from baseline FLAIR MRI scans. The study tested five deep learning models against a population-…
-
New framework generates pseudo-labels for material CT scans without GPUs
Researchers have developed TEMPL, a novel framework for generating pseudo-labels in material CT scans without requiring pre-trained models or GPUs. This model-free approach utilizes thresholding, morphological dilation,…
-
Spatial Lifting technique enhances dense prediction with fewer parameters
Researchers have introduced Spatial Lifting (SL), a new technique for dense prediction tasks that enhances performance while reducing computational costs and model parameters. SL works by transforming standard inputs, l…
-
Synthetic data framework AxonSynth enables zero-shot 3D axon segmentation
Researchers have developed AxonSynth, a novel framework that generates synthetic data for training 3D axon segmentation models. This approach aims to overcome the challenge of obtaining expensive manual annotations for …
-
New GAN model enhances 3D medical image segmentation accuracy
Researchers have developed DE-GAN, a novel generative adversarial network designed to improve 3D medical image segmentation, particularly for brain tumors. This model synthesizes adaptive FLAIR images by incorporating i…
-
AI models tackle PET attenuation correction challenge · 2 sources tracked
Two research papers detail approaches to the Big Cross-Modal Attenuation Correction (BIC-MAC) challenge, focusing on synthesizing pseudo-CT images from PET, MRI, and topogram data. The first paper highlights the importa…
-
New pipeline generates ML-ready datasets for 3D fluid dynamics simulations
Researchers have developed ChannelFlow-Tools, an open-source pipeline designed to generate machine-learning-ready datasets for 3D obstructed channel flows. This configuration-driven system integrates procedural obstacle…
-
New DINO-3DRA framework improves 3D cerebral aneurysm segmentation
Researchers have developed DINO-3DRA, a novel framework for segmenting cerebral aneurysms in 3D rotational angiography (3DRA) data. The system effectively transfers semantic knowledge from 2D foundation models, like DIN…
-
New GPU framework enables nanoscale biological analysis without dense annotations
Researchers have developed a novel GPU-accelerated framework to analyze nanoscale biological structures from anisotropic confocal microscopy data. This method avoids the need for dense volumetric annotations by training…
-
NeuroRefiner system enhances 3D neuron segmentation accuracy
Researchers have developed NeuroRefiner, a novel multi-agent system designed to improve the accuracy of 3D neuron segmentation in fluorescence microscopy. This system mimics the iterative process of human experts by emp…
-
New Benchmark Compares Deep Learning Models for Brain Tumor Segmentation
Researchers have developed a unified benchmark to compare deep learning models for 3D brain tumor segmentation from MRI scans. The study evaluates five state-of-the-art models, including CNNs, Transformer-based models, …
-
New AI model generates 3D dynamic humans from text in under a minute
Researchers have developed 4DHumanDiff, a novel diffusion framework capable of directly generating dynamic 3D human models from text prompts. This method bypasses the need for intermediate video synthesis or per-scene r…
-
AI models for MRI segmentation show limited generalisability despite intensity normalisation
Researchers conducted a systematic benchmark of seven intensity normalization methods for 3D knee MRI segmentation using a 3D U-Net model. The study found that while methods like Z-score, Nyúl histogram matching, and CL…
-
Deep learning model improves tree biomass estimation with continuous data mapping
Researchers have developed a new method for estimating tree biomass distribution using deep learning, shifting from discrete plot-level data to continuous Horizontal Biomass Distribution (HBD) mapping derived from Quant…
-
Render-FM achieves real-time photorealistic CT scan rendering
Researchers have developed Render-FM, a novel feedforward model designed for real-time photorealistic volumetric rendering of CT scans. This model significantly speeds up the rendering process, reducing it from hours or…
-
New AI framework improves trauma detection in CT scans
Researchers have developed CT-VDETR, a novel framework for detecting traumatic injuries in CT scans, addressing the challenge of limited voxel-level annotations. The system combines self-supervised pretraining using Mas…
-
New Diffusion Model Enhances Synthesis of MS Lesion MRI Scans
Researchers have developed Lesion-DDPM, a novel 3D conditional diffusion framework designed to synthesize medical images for multiple sclerosis (MS) research. This method specifically enhances the generation of images t…
-
New ViASNet model predicts viewer engagement in video ads
Researchers have developed ViASNet, a novel deep saliency prediction model designed for short-form video advertising. This model, based on the 3D U-Net architecture, incorporates audio cues and semantic scene meaning to…
-
AI speeds up cardiac MRI reconstruction, improving image quality
Researchers have developed a novel method for rapid online reconstruction of real-time simultaneous multi-slice (RT-SMS) bSSFP cardiac MRI. This technique utilizes a 3D U-Net deep learning model for artifact suppression…