New method separates mixed distributions without requiring samples of each source.
problem Separating mixed distributions in machine learning and signal processing.
method Neural Egg Separation method iteratively learns to separate known from unknown distributions.
result Neural Egg Separation outperforms current methods in audio and image separation tasks.
This paper addresses converting speech to EGG signals without hardware, improving accuracy.
problem Estimating EGG signals from speech without hardware.
method Optimization of evidence lower bound with KL-divergence minimization.
result The method generates EGG signals that agree with gold standard and outperforms state-of-the-art.
This paper proposes a novel selective autoencoder approach within the framework of deep convolutional networks. The crux of the idea is to train a deep convolutional autoencoder to suppress undesired parts of an image frame while allowing the desired parts resulting in efficient object detection. The efficacy of the fr…
We discuss the structure of "exceptional generalised geometry" (EGG), an extension of Hitchin's generalised geometry that provides a unified geometrical description of backgrounds in eleven-dimensional supergravity. On a d-dimensional background, as first described by Hull, the action of the generalised geometrical O(d…
LGD breaks the chicken-and-egg loop in adaptive SGD by using LSH sampling.
problem Challenging per-iteration cost of adaptive gradient sampling.
method Locality Sensitive Hashing (LSH) sampled Stochastic Gradient Descent (LGD).
result Superior and faster gradient estimation with similar per-iteration cost.
Paper solves the chicken-and-egg problem in unsupervised learning of signal models.
problem Learning signal models from incomplete data when the model is unknown.
method Necessary and sufficient sensing conditions for learning signal models from multiple measurement operators or group invariance.
result Agrees with the fundamental limitations of learning from incomplete data.
Paper identifies resting positions using EGG, ECG, respiration rate, and SpO2.
problem Identifying the resting position for health monitoring.
method Hybrid stacked ensemble machine learning model combining Decision tree, Random Forest, and Xgboost.
result 100% accurate prediction of resting positions.
Improves deep neural networks using soft labels through alternating minimization.
problem Improving deep neural networks training with soft labels.
method Co-Learns DNNs and soft labels via Alternating Minimization of two objectives.
result COLAM achieves improved performance on many tasks with better testing classification accuracy.
DANCE optimizes neural network and accelerator design for faster, more efficient DNN execution.
problem Challenges in optimizing neural network and accelerator design for efficient DNN execution.
method Differentiable approach to co-exploration of accelerator and network architecture design.
result Significantly shorter time to achieve superior accuracy and hardware cost metrics.
New meta-RL method avoids exploration-exploitation trade-off.
problem Learning to explore and exploit simultaneously in meta-RL.
method Developed new objectives for exploration and exploitation.
result DREAM outperforms existing methods on complex tasks.
Neural networks can separate non-separable data using feature maps.
problem Non-separable data in neural networks.
method Characterization of feedforward neural networks and use of feature maps.
result ReLU neural networks can separate concentric data.
Shallow neural nets classify objects perfectly if their distribution is linearly separable.
problem Designing efficient neural networks for classification.
method Constructed shallow sigmoid-type neural networks.
result Achieves 100% accuracy for datasets following a linear separability condition.
We present a modification of the so-called Parrondo's paradox where one is allowed to choose in each turn the game that a large number of individuals play. It turns out that, by choosing the game which gives the highest average earnings at each step, one ends up with systematic loses, whereas a periodic or random seque…
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.
Quantifying behavior is crucial for many applications in neuroscience. Videography provides easy methods for the observation and recording of animal behavior in diverse settings, yet extracting particular aspects of a behavior for further analysis can be highly time consuming. In motor control studies, humans or other …
DSI measures dataset separability for neural networks.
problem Difficulty in separating different classes of data in neural networks.
method Created the Distance-based Separability Index (DSI) to quantify dataset separability.
result DSI effectively measures dataset separability and indicates similar distributions of different classes.
The Midscribability Theorem, which was first proved by O. Schramm, states that: given a strictly convex body K⊂R3 with smooth boundary and a convex polyhedron P, there exists a polyhedron Q⊂RP3 combinatorially equivalent to P which midscribes K. Here the word "midscribe" me…
Spectral analysis shows neural networks separate from linear methods in approximating functions.
problem Separating two-layer neural networks from linear methods in function approximation.
method Spectral-based approach using Kolmogorov width and kernel spectrum.
result Upper and lower bounds on separation, explicit hard functions identified.
Boosts neural network performance by improving weight separability.
problem Improving the separability of weight vectors in neural networks.
method Proposes a new evaluation metric and feed-backward reconstruction loss to encourage weight separability.
result Improves visual recognition performance across various tasks.
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.
Proposes a two-step method for sound source separation.
problem Improving sound source separation performance.
method First, learn a latent space transform. Second, train a separation module in the latent space.
result The proposed method achieves better performance than joint learning approaches.
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.
Randomly initialized neural networks can linearly separate arbitrary sets.
problem Mapping two arbitrary sets to linearly separable sets.
method Randomly initialized one-layer neural networks with sufficient width.
result With high probability, these networks can transform two sets into linearly separable sets.
Constructs classifiers for neural networks with specific data configurations.
problem Finding global minima of deep ReLU neural networks on sequentially separable data.
method Explicitly constructs zero loss neural network classifiers using cumulative parameters and truncation maps.
result Global minimizers can be described with a limited number of parameters based on the data structure.
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.
Adversarial noises are linearly separable for random neural networks.
problem The challenge of adversarial examples in neural networks.
method Theoretical proof and empirical evidence for two-layer networks with random initialization and neural tangent kernel setup.
result Adversarial noises are linearly separable with corresponding labels.
Paper shows how SFA fits into FBM framework for time series separation.
problem Identifying time series decomposition in flow-based models.
method Combining SFA and FBM to make time series decomposition identifiable.
result Time series decomposition becomes identifiable using SFA and FBM.
A new unsupervised method separates speech sources without requiring labeled data.
problem Lack of supervised data for effective neural source separation.
method Uses a complex Gaussian mixture model (cGMM) for joint training of separation and localization networks.
result The method outperforms conventional initialization methods in monaural and multichannel separation.
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.
Spatial blind source separation simplifies multivariate spatial prediction.
problem Predicting multivariate measurements at unobserved locations with spatial dependencies.
method Spatial blind source separation as a pre-processing tool compared to Cokriging and neural networks.
result Spatial blind source separation simplifies spatial prediction by avoiding cross-dependencies.
New method separates objects from images using deep neural networks trained to inpaint.
problem Fully self-supervised instance separation of occluded objects in images.
method Maximizes independence of two image regions given a fully self-supervised inpainting network.
result Method achieves similar segmentation performance to fully supervised methods on microscopy image datasets.
New findings on depth vs. width in neural networks, showing depth can improve learnability.
problem Understanding the role of depth in neural networks, especially when width is unbounded.
method Analyzing sample complexity for learnability in norm-controlled depth-2 and depth-3 ReLU networks.
result Depth can improve learnability of functions that are otherwise unlearnable with depth-2 networks.
New method separates multiple voices in mixed audio.
problem Separating multiple simultaneous speakers in audio.
method Gated neural networks trained at multiple steps, selecting actual number of speakers.
result Outperforms current state of the art for more than two speakers.
Study shows post-COVID commodity futures returns and volatility changed for different products.
problem Analyzing how the pandemic affected Chinese commodity futures markets.
method Empirical analysis of commodity futures returns and cointegration before and after the pandemic.
result Post-COVID, some commodity futures returns increased significantly, while others saw higher volatility.
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.
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.
Paper tackles policy selection in offline RL without hyperparameters.
problem Selecting between policies and value functions in offline RL.
method Designs hyperparameter-free algorithms based on BVFT for policy selection.
result Demonstrates effectiveness in discrete-action benchmarks like Atari.
Gradient descent converges to perfect classification in neural nets for non-separable data.
problem Classifying linearly non-separable data using neural networks.
method Analysis of gradient descent dynamics in neural networks with sufficient but not large number of neurons.
result Gradient descent converges to global minima with perfect classification in the landscape of minimization problems.
Law explains how deep networks separate data for classification.
problem Black-box nature of deep learning limits architecture design and interpretation.
method Studied how deep neural networks process data in intermediate layers.
result Law of geometric data separation emerges in various architectures and datasets.
A deep clustering model learns to separate audio sources without supervision.
problem Training deep clustering models requires supervision, limiting their applicability.
method Proposes an unsupervised spatial clustering approach to train a deep clustering system.
result The deep clustering model achieves similar performance to a multi-channel teacher without supervision.
Deep learning model separates syntax and semantics for better language generalization.
problem Standard deep learning methods struggle with systematic generalization in natural language.
method Implemented a Syntactic Attention model that separates syntactic and semantic processing.
result The Syntactic Attention model outperforms standard methods on a compositional generalization task.
Neural networks can learn Boolean circuits with local correlation.
problem Learning Boolean circuits with neural networks is computationally hard.
method Observing local correlation between input patterns and target labels, focusing on tree-structured Boolean circuits.
result Local correlation determines the success or failure of optimization in learning Boolean circuits.
Proves depth 2 neural networks can't approximate certain functions as well as depth 3 networks.
problem Approximating functions with depth 2 networks in high dimensions.
method Lower bound proof using worst-to-average-case random self-reducibility.
result Proves depth 2 networks can't approximate certain functions as well as depth 3 networks, resolving an open problem.
Study on size and depth of neural networks for approximating benign functions, showing barriers and explicit results.
problem Understanding how size and depth of neural networks affect their ability to approximate benign functions.
method Analyzing ReLU networks for benign functions, proving barriers and explicit results.
result Explicit benign functions that cannot be approximated by networks of certain sizes or depths, showing barriers to size and depth separation.
SepIt improves speech separation for multiple speakers.
problem Improving speech separation for multiple speakers in single channel recordings.
method SepIt uses a deep neural network that iteratively improves estimates of different speakers based on mutual information.
result SepIt outperforms state-of-the-art methods for 2, 3, 5, and 10 speakers.
Analyzes memory time span in LSTMs for multi-speaker speech separation.
problem Understanding how long-term dependencies are handled by LSTMs in speech separation tasks.
method Leaked state variable with controlled lifetime to evaluate task performance.
result Estimates the time span LSTMs exploit in multi-speaker speech separation.
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.
New algorithm proves deep networks can learn better than shallow ones.
problem Understanding the power difference between shallow and deep neural networks.
method Identifying a class of Boolean functions and proving that logarithmic-depth networks can learn them efficiently using hierarchical reconstruction.
result First algorithmic separation between constant-depth and logarithmic-depth neural networks.