AC/DC
PulseAugur coverage of AC/DC — every cluster mentioning AC/DC across labs, papers, and developer communities, ranked by signal.
- 2026-06-15 product_launch Renault Group's ACDC Hangzhou R&D office has officially opened, focusing on AI and software development. source
3 day(s) with sentiment data
-
New AI models synthesize full cardiac MRI from single frame, ditching ECG gating
Two research papers introduce novel generative frameworks, PhaseFlow3D and PhaseFlow, for synthesizing complete 4D cardiac cine MRI sequences from a single end-diastolic 3D volume. These methods eliminate the need for e…
-
New AI framework reconstructs 3D heart geometry from sparse MRI data
Researchers have developed a novel framework called Local2Mesh for reconstructing 3D left ventricular geometry from sparse 2D cardiac MRI data. This method addresses challenges like inter-slice misalignment and insuffic…
-
Mechanistic Interpretability Research Reveals Objective-Level Recovery Gaps
A new research paper titled "Are We Recovering Mechanisms? Objective-Level Recovery Gaps in Mechanistic Interpretability" has been published on arXiv. The paper investigates the effectiveness of current methods in mecha…
-
New SSS Method Enhances Medical Image Segmentation Using SAM-2
Researchers have developed SSS (Semi-Supervised SAM-2), a novel approach for semi-supervised medical image segmentation that leverages the feature extraction capabilities of SAM-2. This method aims to improve segmentati…
-
New method enhances 4D medical image interpolation with Fourier deformation
Researchers have developed a novel method for 4D medical image interpolation, a technique crucial for analyzing dynamic anatomical changes in applications like cardiac MRI and thoracic CT. The new approach, termed Phase…
-
New method uses low-rank velocity fields for unsupervised 4D medical image interpolation
Researchers have developed a novel method for unsupervised 4D medical image interpolation, which synthesizes intermediate volumes from sparsely sampled sequences. This technique utilizes low-rank velocity fields as a st…
-
New AI framework synthesizes cardiac motion from single scan
Researchers have developed a novel framework for synthesizing full-cycle biventricular motion from a single end-diastolic (ED) anatomy scan. This method addresses the challenge of localized and phenotype-dependent cardi…
-
Graph learning framework tackles transformer circuit localization
Researchers have introduced Graph Circuit Learning (GCL), a novel framework that treats circuit localization in transformer models as a graph machine learning problem. This approach trains a graph neural network (GNN) a…
-
New CDGC-Net advances 3D medical image segmentation accuracy
Researchers have developed CDGC-Net, a novel network designed for 3D medical image segmentation. This network integrates cooperative dual-scale spatial attention with grouped hierarchical channel modeling to effectively…
-
New framework improves few-shot medical image segmentation
Researchers have developed a novel bi-level collaborative learning framework to address the challenges of few-shot scribble-supervised medical image segmentation. This approach utilizes a learnable superpixel model to p…
-
OFD-Net: Teacher-Free Medical Image Segmentation Framework Unveiled
Researchers have developed OFD-Net, a novel framework for semi-supervised medical image segmentation that does not require a teacher network. This approach utilizes an Orthogonal Feature Disentanglement Module (OFDM) to…
-
Biologically inspired preprocessing enhances AI model robustness to lighting shifts
Researchers have developed a novel preprocessing module inspired by the human retina to enhance the robustness of deep neural networks against distribution shifts like varying illumination and weather conditions. This m…
-
SegWithU framework enhances medical image segmentation with uncertainty estimation
Researchers have developed SegWithU, a novel framework for medical image segmentation that accurately estimates uncertainty without requiring multiple inference passes. This post-hoc method augments existing segmentatio…
-
New RF-HiT model offers efficient medical image segmentation
Researchers have developed RF-HiT, a novel Rectified Flow Hierarchical Transformer designed for efficient and accurate medical image segmentation. This model addresses the computational complexity and latency issues of …
-
Metadata-driven pre-training boosts cardiac MRI models
Researchers have developed MetaCLIP-CMR, a novel framework for pre-training cardiac MRI foundation models by leveraging structured acquisition metadata. This approach converts imaging modality, anatomical view, scanner …
-
AI research optimizes camera sensors for autonomous driving segmentation
Researchers have developed a method for co-designing camera sensors and AI models for autonomous driving, focusing on optimizing the sensor's color filter array (CFA) weights. This approach demonstrated significant impr…
-
Neural Phase Correlation framework generalizes image transformation analysis
Researchers have developed a novel framework called Neural Phase Correlation, which generalizes the traditional phase correlation method. This new approach learns a basis for transformations, enabling it to handle dense…
-
Chinese Stock Markets Surge Past 2.5 Trillion Yuan Amidst AI Developments
The Shanghai and Shenzhen stock markets have seen significant trading volume, exceeding 2.5 trillion yuan at one point. The ChiNext index has also reached new historical highs, indicating strong market performance. Amid…
-
Renault Group opens AI and software R&D office in Hangzhou
Renault Group has announced the official opening of its ACDC research and development office in Hangzhou. This new facility will concentrate on cutting-edge technologies such as software, artificial intelligence, and us…
-
New framework maps medical image dataset needs to segmentation model design
Researchers have introduced the Medical Segmentation Dataset Knowledge Card (MS-DKC) framework to better understand the specific requirements of medical imaging datasets for segmentation models. This framework explicitl…