Proposes a new interpretation of separable convolutions.
problem Lack of a thorough explanation for the efficacy of separable convolutions.
method Hybrid interpretation combining depthwise and pointwise convolutions.
result Proposes a new model for understanding separable convolutions.
DSGC unifies graph and grid convolutions.
problem Lack of understanding between graph and grid convolutions.
method Depthwise separable graph convolution.
result DSGC outperforms existing methods on benchmark datasets.
A new method speeds up spectral normalization for neural nets.
problem Efficiently controlling the spectral norm of convolutional layers.
method Depthwise separable convolutions with spectral normalization.
result Significant reduction in computational and memory costs.
Efficient neural network for audio source separation.
problem End-to-end general purpose audio source separation.
method SuDoRMRF structure with simple one-dimensional convolutions for feature aggregation.
result SuDoRMRF achieves high quality audio source separation with minimal computational resources.
Paper presents a provably correct algorithm for CNMF under separable conditions.
problem Convolutive nonnegative matrix factorization (CNMF) under separable assumptions.
method Algorithm exploiting NMF model and existing separable NMF algorithms.
result Guaranteed solution in low noise settings, runs in polynomial time.
Graph convolution improves linear separability and generalizes to out-of-distribution data.
problem Improving linear separability in semi-supervised classification.
method Applying graph convolution to mixtures of Gaussians in a stochastic block model.
result Graph convolution extends the linear separability regime by a factor of 1 / D 1/\sqrt{D} 1/ D . A new neural network separates vocals from music accompaniment.
problem Separating vocals from music accompaniment in recordings.
method Self-attention convolutional neural network (CNN) with densely-connected blocks.
result 19.5% relative improvement in vocals separation.
A new text classification method using separable convolution reduces memory consumption.
problem Manual categorization of documents is inefficient and resource-intensive.
method Introducing a new architecture based on separable convolution for text classification.
result Achieved a drastic reduction in trainable parameters without compromising accuracy.
Improves audio source separation using dilated convolutions and dense connections.
problem Optimizing feature extraction in audio source separation models.
method Adaptive dilated convolutions and dense connections in U-Net architecture.
result Improved performance on MUSDB test dataset.
DSGC improves graph representation learning by modeling object links and attribute relations.
problem Limited modeling capability of existing GCN variants on noisy and sparse real-world networks.
method Dimensionwise separable 2-D graph convolution (DSGC) that filters node features.
result DSGC achieves significant performance gain over state-of-the-art methods for node classification and clustering.
A new neural network separates singing voices more effectively.
problem Separating singing voices from mixed signals with high accuracy.
method MBR-FCN that processes different frequency bands with varying resolutions and filters.
result The MBR-FCN achieves better performance with fewer parameters.
A modified VDCNN model reduces size and latency for mobile platforms.
problem Memory and processing constraints on mobile platforms.
method Temporal Depthwise Separable Convolutions and Global Average Pooling.
result The squeezed model (SVDCNN) is 10x-20x smaller with minimal accuracy loss.
Paper improves speech separation by using deep neural networks for more accurate density priors.
problem Improving the accuracy of source priors for independent vector analysis in speech separation.
method Estimating the derivative of speech density using deep neural networks to optimize performance indices.
result Neural network density priors outperform previous ones in convergence speed and SIR.
CRsAE auto-encoder recovers convolutional dictionary from noisy signals.
problem Recovering a convolutional dictionary from noisy signals.
method Constrained recurrent sparse auto-encoder (CRsAE) architecture.
result CRsAE successfully recovers the underlying dictionary in the presence of noise.
Stanza separates convolutional and fully connected layers for faster deep learning training.
problem Heavy data transfer between workers and servers in distributed deep learning.
method Layer separation: most nodes train convolutional layers, others train fully connected layers only.
result Significant acceleration of training time (1.34x--13.9x) over current systems.
XceptionTime improves hand gesture recognition accuracy using novel deep learning.
problem Improving hand gesture recognition from sparse sEMG signals.
method Depthwise separable convolutions, adaptive pooling, non-linear normalization.
result Significantly improved accuracy (5.71% improvement) in hand gesture recognition.
This paper explains a mechanism called phase collapse that improves image classification accuracy.
problem Understanding the role of non-linearities and convolutional filters in image classification.
method Demonstrates phase collapse as a mechanism that eliminates spatial variability and linearly separates classes.
result Phase collapse improves classification accuracy, while thresholding operators degrade performance.
New method separates and deconvolves signals from single-channel mixtures.
problem Separating and deconvolving individual sources from a single-channel mixture.
method Synthesizing-decomposition (S-D) approach using GAN for sources and optimization for filters and sources.
result Achieves PSNR improvements over existing methods in various tasks.
Convolutional networks outperform fully-connected ones in certain tasks.
problem Understanding the computational advantage of convolutional networks over fully-connected networks.
method Demonstrated a computational advantage through a specific problem class.
result Convolutional networks can solve certain problems that fully-connected networks cannot, even with gradient descent.
Unsupervised mesh disentanglement separates identity and pose.
problem Geometric disentanglement for 3D deformable models.
method CFAN-VAE architecture using conformal factor and normal features.
result CFAN-VAE achieves state-of-the-art performance on unsupervised geometric disentanglement.
Deep learning models improve sound separation across various types of sounds.
problem Developing a universal method to separate arbitrary sounds of different types.
method Created a dataset of mixtures containing arbitrary sounds, investigated mask-based separation architectures, and tested different framewise analysis-synthesis bases.
result STFT outperformed learnable bases in universal sound separation tasks.
Deep convolutional networks provide state of the art classifications and regressions results over many high-dimensional problems. We review their architecture, which scatters data with a cascade of linear filter weights and non-linearities. A mathematical framework is introduced to analyze their properties. Computation…
GCNs distinguish graph models based on embeddings, but depth matters.
problem GCNs distinguish between different random graph models.
method Investigated the power of GCNs of varying depths to distinguish between graph models.
result GCNs with logarithmic depth can distinguish certain graphons, but simpler architectures suffice for others.
The study quantifies how many objects can be linearly classified under all views.
problem Understanding the expressivity of group-equivariant representations.
method Generalization of Cover's Function Counting Theorem to quantify separable dichotomies.
result The fraction of separable dichotomies is determined by the fixed space dimension of the group action.
Enhances CNN feature extractors' separation capacity analysis.
problem Understanding the separation capacity of CNNs.
method Extending Cover's function-counting theory, analyzing scattering networks.
result Identifies factors affecting scattering networks' separation capacity.
Structured linear substitutions improve both efficiency and accuracy in neural networks.
problem Improving neural network efficiency and accuracy tradeoff.
method Replacing linear components in pointwise convolutions with structured linear decompositions.
result Structured layers provide Pareto-optimal benefits in efficiency/accuracy.
New CNN architecture separates 'what' and 'where' in neural data.
problem Estimating individual receptive field locations in neural data.
method Sparse readout layer factorizing spatial and feature dimensions.
result Our network outperforms current models in system identification.
CCTV footage of beef cattle segmented into individual instances.
problem Early detection of animal welfare issues in beef cattle.
method Transformed a fully convolutional network into an instance segmentation network.
result Network outputs a single prediction for every animal in CCTV footage.
The paper explores how different patterns of heterophily affect Graph Neural Networks.
problem Understanding the impact of heterophily on Graph Neural Networks.
method Theoretical analysis and experiments with Heterophilous Stochastic Block Models (HSBM).
result The impact of heterophily on classification depends on the Euclidean distance of neighborhood distributions and the averaged node degree.
BiGraphNet generalizes graph neural networks for more efficient operations.
problem Fragmented graph neural network architectures hinder optimization.
method Explicitly separates input and output nodes, enabling new efficient operations.
result BiGraphNet accelerates and scales computations in hierarchical networks.
Compact neural networks improve ECT macromolecule classification speed without sacrificing accuracy.
problem Efficiently classify millions of macromolecules captured by ECT for faster structural analysis.
method Knowledge distillation to compress complex teacher networks into simpler student networks.
result Compressed models reduce parameter count and prediction time while maintaining accuracy.
New algorithm separates audio sources better using alpha-stable distributions.
problem Improving audio source separation using complex distributions.
method Estimating mixtures of alpha-stable distributions using characteristic function matching.
result Better separation performance than Gaussian-based methods.
Corrected graph convolutions improve node classification on graphs.
problem Oversmoothing in graph convolutions degrades performance.
method Theoretical analysis based on CSBM, spectral analysis for k rounds of corrected graph convolutions.
result Corrected graph convolutions can improve node classification performance exponentially.
Improved U-Nets with various intermediate blocks enhance singing voice separation.
problem Improving singing voice separation accuracy using U-Net architectures.
method Implemented and compared U-Nets with different intermediate spectrogram transformation blocks.
result A specific block type achieves state-of-the-art SDR by 0.9 dB.
Convolutional neural networks improve time series forecasting efficiency.
problem Efficiently forecasting multivariate time series with historical data.
method Adapted WaveNet architecture with dilated convolutions and parallel filters.
result Convolutional network outperforms autoregressive and LSTM models.
The paper bridges spectral and spatial graph convolutions, improving model capacity and transferability.
problem Improving graph neural networks by bridging spectral and spatial design.
method Theoretical demonstration and general framework for spectral analysis, new spectral convolutions, and depthwise separable convolutions.
result General framework allows spectral analysis of ConvGNNs, showing their performance and limits, and proposing new spectral convolutions.
EEGNet creates a compact CNN for EEG-based BCIs across various paradigms.
problem Designing a single CNN for EEG-based BCIs that works across different paradigms.
method Use of depthwise and separable convolutions to create an EEG-specific model.
result EEGNet generalizes across paradigms better than current state-of-the-art approaches.
Study evaluates data augmentation methods for prostate cancer detection in MRI.
problem Limited data for prostate cancer detection in MRI.
method Static application of five augmentation techniques (rotation, flip, crop, translation) to MRI dataset.
result Rotation method improved 2D slice-based AUC to 0.85.
Model classifies environment sounds using multiple feature channels and attention mechanisms.
problem Environment sound classification task.
method Multiple feature channels (MFCC, GFCC, CQT, Chromagram) and attention mechanism in a deep CNN.
result Achieves state-of-the-art performance on three benchmark datasets.
The paper analyzes deep convolutional representations and their properties.
problem Understanding the properties of deep convolutional architectures.
method Introducing a multilayer kernel based on convolutional kernel networks and studying the geometry induced by the kernel mapping.
result Characterization of the RKHS and its relation to model complexity and generalization.
ScDCFNet improves multiscale image classification with reduced model size.
problem Improving performance in multiscale image classification.
method Decomposed convolutional filters for ST-equivariant CNNs.
result ScDCFNet achieves significantly improved performance in multiscale image classification.
Improved speech separation and enhancement using neural beamforming.
problem Challenging speech separation and enhancement in reverberant environments.
method Sequential neural beamforming combining spectral and spatial separation methods.
result Average improvement of 2.75 dB in scale-invariant signal-to-noise ratio and 14.2% absolute reduction in speech recognition metric.
Deep audio prior uses neural networks to solve audio problems without data.
problem Challenging audio problems like source separation, editing, and synthesis.
method Randomly-initialized neural network with carefully designed audio prior.
result Superior audio results on Universal-150 benchmark dataset.
GCNs improve multi-layer network classification by expanding the distance between means.
problem Improving multi-layer network classification with graphical information.
method Theoretical and empirical study of graph convolutions in multi-layer networks.
result Graph convolutions expand the classification regime by a factor of 1 / E m d e g 4 1/\sqrt[4]{\mathbb{E}{
m deg}} 1/ 4 E m d e g . ANTs integrate neural networks and decision trees for better performance.
problem Combining neural networks and decision trees for improved performance.
method Adaptive Neural Trees (ANTs) integrate representation learning into decision tree structures, using backpropagation for adaptive growth.
result ANTs achieve competitive performance on classification and regression datasets, with benefits in lightweight inference, feature separation, and adaptable architecture.
Convolutional neural networks outperform other architectures in streaming time series classification.
problem Efficient deep learning models for real-time data streams.
method Asynchronous dual-pipeline deep learning framework for real-time predictions.
result Convolutional architectures achieve higher accuracy and efficiency in streaming time series classification.
SVM with local features improves human action recognition.
problem Improving human action recognition in videos.
method Local appearance and motion features extracted using CNNs, concatenated, and used with SVM for classification.
result SVM with local features outperforms previous methods on benchmark datasets.
Study reveals how neural network architectures bias their learning based on feature directions.
problem Understanding how neural network architectures bias their learning based on feature directions.
method Defined neural anisotropy directions (NADs) to encapsulate the directional inductive bias of architectures and provided an efficient method to identify them.
result NADs characterize the features used by CNNs to discriminate between different classes for the CIFAR-10 dataset.