Extracts object-centric frames from unlabeled images.
problem Extracting abstract models of 3D objects from visual measurements.
method Viewpoint factorization and dense equivariant labelling neural network.
result Extracts dense object-centric coordinate frames invariant to deformations.
Proves unique maps from certain spaces to others.
problem Uniqueness of equivariant harmonic maps into specific spaces.
method Analyzes maps into irreducible symmetric spaces and Euclidean buildings.
result Proves uniqueness of maps for certain actions.
Generic density of equivariant min-max hypersurfaces in Riemannian manifolds.
problem Finding generic density of equivariant min-max hypersurfaces in Riemannian manifolds.
method Weyl asymptotic law for G-equivariant volume spectrum, generic density result. result Generic density of equivariant min-max hypersurfaces in Riemannian manifolds.
Problems of dense and closed extension of actions of compact transformation groups are solved. The method developed in the paper is applied to problems of extension of equivariant maps and of construction of equivariant compactifications.
The paper develops methods for calculating equivariant homology from Morse functions.
problem Calculating equivariant homology from equivariant Morse functions.
method Alter equivariant Morse functions to stable ones, use generic equivariant metrics, and analyze the Morse spectral sequence.
result Equivariant Morse functions induce a filtration that computes equivariant homology.
Automates galaxy morphology classification with less human labelling.
problem Insufficient human-labeled galaxy images for accurate classification.
method Developed a VAE with equivariant transformer layers and a classifier network.
result Improves accuracy with fewer labels and unlabelled data.
Constructs equivariant embeddings of Hermitian symmetric spaces into tangent spaces.
problem Embedding Hermitian symmetric spaces into their tangent spaces.
method Using polarity of the K-action to construct equivariant embeddings.
result Characterizes holomorphic/symplectic embeddings and realizes submanifolds.
New techniques enforce sparseness in recurrent models, reducing memory usage.
problem Reducing memory usage in recurrent sequence models for NLP.
method Enforcing sparseness upfront in recurrent layers for language modeling and sequence labeling.
result Predefined sparseness leads to similar performance with fewer parameters.
New method tracks all bees in a hive with high accuracy.
problem Efficient tracking of multiple objects in dense configurations.
method Combining CNNs with U-Net architecture and temporal regularities.
result Near human-level performance with reduced network size.
Let L be a Lie group and Lambda a lattice in L. Suppose G is a non-compact simple Lie group realized as a Lie subgroup of L, and the image of G on L/Lambda is dense. Let c be a diagonalizable element of G not contained in a compact subgroup. Let U be the expanding horospherical subgroup of G associated to c. Let Omega …
A scalable graph-based SSL method for large-scale data with few labels.
problem Challenges in semi-supervised learning with limited labeled data and large unlabeled data.
method Constructs a graph from a small set of high-dense vertexes to learn relationships and improve performance.
result Achieves good classification performance, especially with few labels.
Paper tackles rDR classification and lesion segmentation using self-supervised equivariant learning and attention-based MIL.
problem Classifying rDR and segmenting lesions from image-level labels.
method Integrates self-supervised equivariant attention mechanism (SEAM) with attention-based multi-instance learning (MIL).
result Achieved AU ROC of 0.958 on Eyepacs dataset, outperforming state-of-the-art.
PRISM integrates diverse rewards in MORL, improving sample efficiency and Pareto coverage.
problem Heterogeneous MORL where dense objectives dominate, leading to poor sample efficiency.
method PRISM uses reflectional symmetry and ReSymNet to reconcile temporal-frequency mismatches and accelerate exploration.
result PRISM consistently outperforms sparse-reward baselines and oracles, achieving significant Pareto gains.
We prove an implicit function theorem for functions on infinite-dimensional Banach manifolds, invariant under the (local) action of a finite dimensional Lie group. Motivated by some geometric variational problems, we consider group actions that are not necessarily differentiable everywhere, but only on some dense subse…
Adaptive sampling detects local concept drift with limited labels.
problem Detecting local concept drift in dynamic environments with scarce labels.
method Combines residual-based exploration and exploitation with EWMA monitoring.
result Superior performance in label efficiency and drift detection accuracy.
The paper introduces models to learn generalized transformation equivariant representations.
problem Capturing intrinsic visual structures equivariant to various transformations.
method Deterministic and probabilistic AutoEncoding Transformations (AET and AVT) models trained to learn visual representations from generic groups of transformations.
result Generalized TERs (GTERs) that are equivariant to transformations in a more general fashion.
This paper constructs and proves the uniqueness of pluriharmonic maps to Euclidean buildings.
problem Existence and uniqueness of pluriharmonic maps to Euclidean buildings.
method Constructs a ρ-equivariant pluriharmonic map with specific asymptotic behavior and proves its uniqueness.
result Uniqueness of pluriharmonic maps to Euclidean buildings.
New method learns robot actions from videos without explicit labels.
problem Training robots to perform tasks from few demonstrations.
method Uses images and text for task-agnostic and general representation, synthesizes hallucinated actions, and applies dense correspondences.
result Trains robot policies solely from RGB videos, achieving diverse tasks across different robots and environments.
New methods improve translation-equivariant neural processes for modeling unknown functions.
problem Modeling unknown latent functions from irregularly sampled measurements.
method Volterra series and set Fourier convolutions to address translation-equivariance and efficiency.
result Improved translation-equivariant neural processes with analytical transparency and linear scalability.
We present a method for training multi-label, massively multi-class image classification models, that is faster and more accurate than supervision via a sigmoid cross-entropy loss (logistic regression). Our method consists in embedding high-dimensional sparse labels onto a lower-dimensional dense sphere of unit-normed …
New graph foundation models respect symmetries for broader applicability.
problem Tailored graph machine learning architectures limit broader applicability.
method Investigates symmetries for label and feature permutations, proving network universal approximator.
result Universal approximator on multisets respecting node and feature permutations.
Classifies actions of tori on manifolds up to diffeomorphisms.
problem Classifying actions of tori on manifolds up to diffeomorphisms.
method Using triples (Q, λ, c) to classify actions, where Q is a manifold-with-corners, λ is a unimodular labelling, and c is a cohomology class.
result Classifies locally standard smooth actions of T up to equivariant diffeomorphisms.
RIO uses rotation-equivariance to train robust inertial odometry models.
problem Training robust inertial odometry models with limited labeled data.
method Rotation-equivariance as self-supervisor, adaptive Test-Time Training (TTT).
result RIO-trained models achieve on-par performance with full-labeled data models.
APLC-XLNet improves XMTC by clustering labels and reducing computational time.
problem Efficiently tagging texts with many labels from a large set.
method Fine-tunes XLNet with APLC to approximate cross entropy loss.
result Achieved state-of-the-art results on XMTC benchmarks.
Develops a new calculus for studying operators on principal bundles.
problem Investigates G-equivariant operators on principal bundles over manifolds. method Introduces Borel-Weil calculus to analyze G-equivariant (pseudo)differential operators. result Explicit conditions for rapid mixing in dynamical systems and spectral theory results for sub-elliptic Laplacians.
Metric evaluates symmetry-breaking in datasets, revealing severe biases.
problem Symmetry-breaking in datasets can hinder the performance of symmetry-aware methods.
method Developed a metric to quantify symmetry-breaking using a two-sample classifier test.
result Symmetry-breaking can prevent optimal performance of invariant methods, even when labels are invariant.
New optimizer designs respect symmetry, improving deep learning models.
problem Optimizers lack respect for symmetry in neural networks.
method Introduce symmetry-compatible principle for optimizer design.
result Symmetry-compatible optimizers improve model performance.
LaMP neural networks model label interactions for multi-label classification.
problem Efficiently modeling label interactions in multi-label classification.
method Label Message Passing (LaMP) Neural Networks, treating labels as nodes on a graph, compute hidden representations conditioned on input using attention-based message passing.
result Significantly outperforms state-of-the-art multi-label classification models on seven real-world datasets.
SOLAR improves search efficiency and accuracy with sparse, orthogonal embeddings.
problem Bottleneck of indexing large dense vectors and NNS for query efficiency and accuracy.
method Proposes SOLAR embeddings: sparse, orthogonal, learned, and random vectors across multiple GPUs.
result Successfully trains 500K dimensional SOLAR embeddings for 1.6M books and multi-label classification.
LangDA improves domain adaptation for semantic segmentation by learning context-aware scene descriptions.
problem Improving domain adaptation for semantic segmentation with dense prediction tasks.
method LangDA learns contextual relationships between objects via VLM-generated scene descriptions and aligns image features with text representation.
result LangDA sets new state-of-the-art across three DASS benchmarks, outperforming existing methods.
The paper describes a parametrisation of harmonic maps from a 2-torus to the 3-sphere.
problem Understanding the moduli space of equivariant harmonic maps from a 2-torus to the 3-sphere.
method Explicit parametrisation using spectral data and line bundles.
result The space of spectral data is a fibre bundle over the space of spectral curves, with nontrivial structure for certain invariance groups.
New method finds sparse networks without labels, improving performance.
problem Sparse connectivity in neural networks to reduce memory and energy demands.
method Neural Tangent Transfer method to find sparse networks without labels.
result Sparse networks achieve higher classification performance and faster convergence.
New neural network architectures use signed permutation representations for finite groups, improving performance.
problem Designing and optimizing neural networks for finite groups with signed permutation representations.
method Introduces G-invariant deep neural networks with densely connected layers and signed permutation representations. result Signed permutation representations lead to significantly better performance in classification tasks.
Enhances GCNs using VAT for better node classification.
problem Limited use of unlabeled data in GCNs.
method Virtual Adversarial Training (VAT) on GCN supervised loss.
result Improves GCN generalization performance.
The aim of this paper is to classify simply connected 6-dimensional torus manifolds with vanishing odd degree cohomology. It is shown that there is a one-to-one correspondence between equivariant diffeomorphism types of these manifolds and 3-valent labelled graphs, called torus graphs introduced by Maeda-Masuda-Panov. …
This note proves equivariant de Rham cohomology for quotient spaces.
problem Computing de Rham cohomology of quotient spaces under group actions.
method Equivariant identification of de Rham complexes using foliation theory.
result Canonical isomorphism of de Rham complexes for quotient spaces.
ProbE model improves relational implication detection to 0.8143.
problem Improving inference of relational data to extract more useful information.
method Formal probabilistic model of relational implication using estimators based on empirical distribution.
result ProbE model outperforms existing approaches, achieving 0.8143 AUC on evaluation dataset.
Develops methods for learning similarity metrics and group-equivariant representations.
problem Learning discriminative representations for comparing objects, especially when limited computational resources are available.
method Proposes new formulations for metric learning, including extensions for kNN regression and asymmetric similarity learning. Introduces a computationally inexpensive approach for estimating metrics using gradient estimates. Develops SO(3)-equivariant neural networks for spherical data.
result Demonstrates improved k-NN accuracy and regression performance through novel metric learning formulations.
ARMA nets expand receptive fields for dense prediction tasks.
problem Global information in dense prediction problems is challenging for traditional convolutional layers.
method ARMA layers with adjustable autoregressive coefficients replace traditional convolutions.
result ARMA networks improve dense prediction tasks including video prediction and semantic segmentation.
WiGS improves active learning for regression by dynamically selecting informative samples.
problem Reducing labeling costs in regression tasks.
method Formulated as a reinforcement learning problem, WiGS adapts the exploration-investigation balance.
result WiGS outperforms static methods in accuracy and labeling efficiency, especially in irregular data density.
This work tackles out-of-distribution detection using multiple semantic label representations.
problem Detecting neural networks' performance on out-of-distribution examples.
method Using multiple semantic dense representations instead of sparse representation as target labels.
result The proposed method compares favorably with previous work on out-of-distribution detection.
We propose Rademacher complexity bounds for multiclass classifiers trained with a two-step semi-supervised model. In the first step, the algorithm partitions the partially labeled data and then identifies dense clusters containing κ predominant classes using the labeled training examples such that the proportion of t…
SCENE-Net improves 3D point cloud segmentation with low resource usage and transparency.
problem Lack of resources and transparency in 3D semantic segmentation models.
method SCENE-Net uses signature shapes identified via GENEOs to achieve semantic segmentation with minimal resources.
result SCENE-Net achieves comparable IoU to state-of-the-art methods with less data and computational resources.
A new deep learning method for tissue-cleared image registration.
problem Efficient registration of high-resolution tissue-cleared images.
method Densely connected convolutional architecture for deformable image registration, unsupervised training.
result Comparable and superior registration performance to state-of-the-art methods, especially at higher resolutions.
Extends RT TQFT to include surface defects and line defects.
problem Distinguishing non-isotopic embeddings of surfaces in 3D bordisms.
method Uses modular tensor categories, symmetric Frobenius algebras, and equivariant multi-modules.
result Invariant distinguishes non-isotopic embeddings of 2-tori but not 2-spheres.
A novel tracking method for dense honeybee colonies using pixel personality.
problem Tracking large numbers of densely-arranged, interacting objects in a 2D environment.
method Segmentation-based object detection followed by adaptive object recognition through visual appearance.
result Reconstructed ~46% of trajectories in 5 minutes and 71% of tracks for at least 2 minutes.
Optimizes submodular extensions for efficient marginal estimation.
problem Efficiently compute approximate marginals for submodular energy functions.
method Equivalence between submodular extensions and LP relaxations for MAP estimation; worst-case optimality established.
result Worst-case optimal submodular extension for various models.
ECN framework improves training on noisy structured labels.
problem Structured errors in fine-grained annotations lead to biased models.
method Error-Correcting Networks (ECN) framework.
result ECN improves fine-grained annotation prediction.