rotation group SO(3)
PulseAugur coverage of rotation group SO(3) — every cluster mentioning rotation group SO(3) across labs, papers, and developer communities, ranked by signal.
7 day(s) with sentiment data
-
New neural ODE method discovers Lie symmetries and first integrals from data
Researchers have developed LatentLieODE, a novel neural ordinary differential equation (ODE) method capable of identifying linear Lie symmetries and first integrals from observed time-series data, such as videos. This m…
-
New method CanDO recovers performance in rotated point cloud models
Researchers have developed CanDO, a novel method for improving the performance of black-box point cloud models that have been degraded by test-time rotations. Unlike previous approaches that require model weights or gra…
-
New SO(3)-equivariant CNN advances spherical data analysis
Researchers have developed a novel SO(3)-equivariant convolutional neural network (CNN) designed for data residing on the sphere, specifically utilizing the HEALPix format. This architecture incorporates steerable filte…
-
New algebraic formalism unifies deep learning architectures
Researchers have introduced "Product Interactions," a novel algebraic formalism designed to unify and systematically construct various deep learning architectures. This framework represents data as algebraic elements an…
-
New neural operator model ensures rotation-robust predictions for fluid dynamics
Researchers have developed a new neural network model called the Invariant-Conditioned Isotropic Kernel Neural Operator (IKNO) to address issues with coordinate dependence in neural surrogates for partial differential e…
-
New NC-SSM models overcome limitations of Mamba, S4D, and LRU
Researchers have introduced Non-Commutative State Space Models (NC-SSM) to address limitations in existing selective state space models like Mamba. These new models, including SO(3)-SSM and Cayley-SSM, extend state tran…
-
New descriptor precisely analyzes 3D shapes, preserving chirality
Researchers have developed a new descriptor for analyzing 3D shapes with spherical topology, addressing limitations of existing methods. This descriptor precisely removes dependencies on parameterization, pose, and scal…
-
AI predicts optimal build orientation for 3D-printed dental parts
Researchers have developed a machine learning approach to predict the optimal build orientation for dental parts manufactured using selective laser melting (SLM). By training models on approximately 2400 patient-specifi…
-
AI model enhances crystal orientation map resolution using symmetry
Researchers have developed a novel Symmetry-Group-Aware Super-Resolution Attention Network (SG-SRAN) designed to enhance the resolution of crystal orientation maps. This network uniquely incorporates crystal symmetry an…
-
New research explains why optimizers struggle with equivariant networks
Researchers have identified a key reason why certain optimizers like Muon outperform Adam when training equivariant neural networks. The issue stems from how Adam handles learning rates across different blocks within an…
-
New TRI-HAR Framework Achieves Rotation-Invariant Activity Recognition
Researchers have developed TRI-HAR, a novel framework for Human Activity Recognition (HAR) using wearable IMUs. This system is designed to be robust to independent orientation shifts between IMUs at different body locat…
-
New neural network framework learns complex Lie-Poisson system dynamics
Researchers have developed Latent Lie-Poisson Neural Networks (LLPNNs), a novel framework designed to learn and predict the dynamics of Lie-Poisson systems directly from observable data. These systems are crucial for mo…
-
New SE(3)-MeanFlow method accelerates protein backbone generation
Researchers have developed SE(3)-MeanFlow, a novel generative framework for protein backbone design. This method operates on Lie group geometry, enabling faster and more efficient generation compared to existing diffusi…
-
New frameworks tackle scale mismatch and rotation in point-cloud registration
Two new research papers introduce novel frameworks for point-cloud registration, a critical task in 3D perception for robotics. The first, R-SLPR, addresses the challenge of aligning small or incomplete point clouds wit…
-
New framework analyzes information discarded by ML models
Researchers have developed a new framework to analyze the information discarded by machine learning models when inputs have a Lie group action. This framework quantifies the symmetry invisible to the model by defining a…
-
New SO3UFormer architecture enhances rotation-robustness in panoramic AI models
Researchers have developed SO3UFormer, a novel neural network architecture designed to improve the robustness of panoramic dense-prediction models. Unlike existing models that rely on gravity-aligned assumptions, SO3UFo…
-
New spectral embedding method incorporates group symmetries for improved data analysis
Researchers have developed a new spectral embedding method that incorporates group symmetries, such as rotations, into affinity kernels. This approach improves dimensionality reduction and clustering for datasets with i…
-
Geometric Algebra Layers Show Advantage in Deep 3D Learning
A new study published on arXiv investigates the effectiveness of geometric algebra layers in neural networks for learning 3D vector laws. The research compares Clifford algebra Cl(3,0) primitives against a simpler scala…
-
New framework unifies geometry-preserving neural architectures on manifolds
Researchers have developed a unified framework for geometry-preserving neural architectures, organizing them based on where and how geometric constraints are enforced. This work addresses theoretical gaps by proving app…
-
Diffeomorphic optimization method enhances protein design and data manifold learning
Researchers have introduced a novel method called diffeomorphic optimization, designed to improve the process of optimizing objectives on complex data manifolds. This technique leverages diffusion and flow models to map…