deep neural network
PulseAugur coverage of deep neural network — every cluster mentioning deep neural network across labs, papers, and developer communities, ranked by signal.
9 day(s) with sentiment data
-
New DPG loss functions enhance neural network accuracy for PDEs
Researchers have developed new residual-based loss functions for machine learning models that aim to accurately predict solutions for parameter-dependent partial differential equations (PDEs). These functions, particula…
-
FPGA platform accelerates approximate multiplier evaluation for DNNs
Researchers have developed FAME, a new platform utilizing FPGAs to accelerate the evaluation of approximate multipliers for deep neural networks. This hardware-based approach significantly reduces the time needed to ass…
-
OptiPrime framework optimizes private DNN inference with protocol-hardware co-design
Researchers have developed OptiPrime, a framework designed to improve the efficiency of private deep neural network (DNN) inference. This framework addresses the latency issues associated with hybrid homomorphic encrypt…
-
Deep learning model predicts ocean currents for plastic debris cleanup
Researchers have developed Drift Field Net (DFN), a deep neural network designed to predict ocean surface flow fields using satellite observations. This model aims to improve forecasts of particle drift, crucial for str…
-
NeuroFlex enables lossless ANN-SNN co-execution for efficient sparse inference
Researchers have developed NeuroFlex, a novel accelerator design that allows for the co-execution of artificial neural networks (ANNs) and spiking neural networks (SNNs) at the element level. This approach enables each …
-
Research paper details temperature effects on analog DNN inference
A new research paper explores the impact of temperature on analog deep neural network (DNN) inference, particularly for resource-constrained devices like mobile phones. The study found that temperature significantly deg…
-
VertiFuseX: New LSTM Architecture Boosts Financial Forecasting Accuracy
Researchers have developed VertiFuseX, a novel deep learning architecture designed for more generalizable financial forecasting. This hybrid LSTM model utilizes a unique penultimate-layer vertical fusion of multi-scale …
-
Deep learning model classifies galaxy morphology using crowd-sourced data
Researchers have adapted a deep neural network, specifically a convolutional neural network (CNN), for the morphological classification of galaxies using crowd-sourced annotations from the Galaxy Zoo 1 dataset. The stud…
-
New BROT Method Achieves Optimal Transport Map Estimation in Machine Learning
Researchers have introduced BROT (Barycentric Regression for OT), a novel two-step method for estimating optimal transport (OT) maps, which are crucial for aligning probability distributions in machine learning. This ap…
-
New theory enables linear separability for compact datasets using deep neural networks
A new theoretical framework has been developed for relocating compact sets in n-dimensional space using diffeomorphisms, with potential applications in data classification. The research demonstrates that such collection…
-
New DVFS scheduler eliminates thermal throttling for edge AI on passively cooled hardware
Researchers have developed a new Dynamic Voltage and Frequency Scaling (DVFS) scheduler designed to prevent thermal throttling in passively cooled edge devices running deep neural networks. This scheduler utilizes time-…
-
Deep graph generative models show promise for realistic network simulation
A new paper explores the effectiveness of deep graph generative models in creating realistic synthetic networks for research. By analyzing these models from a network science perspective, the study found that certain de…
-
New framework evolves adaptable arithmetic circuits for AI efficiency
Researchers have developed CircuitsDNA, a novel evolutionary framework designed to automatically create arithmetic circuits that can dynamically adjust their accuracy for efficiency. This system integrates multi-thresho…
-
General Coded Computing framework introduced for straggler-resilient ML · 2 sources tracked
Two new arXiv papers introduce General Coded Computing (GCC), a framework designed to improve the efficiency and resilience of distributed computing, particularly for machine learning tasks. The first paper proposes a m…
-
Soft EMG interface enables 97.2% accurate silent speech recognition
Researchers have developed a novel soft electromyography (EMG) interface for silent speech recognition (SSR) that can be worn on the hand. This device uses a fingertip electrode positioned near the lips to capture EMG s…
-
New $(\text{DNN})^2$ method enhances neural network verification
Researchers have developed a new method called $(\text{DNN})^2$ to improve the verification of deep neural networks, particularly those using rectified linear units (ReLUs). Existing methods often provide overly conserv…
-
New framework improves neural decoding by aligning with intermediate DNN representations
Researchers have developed a new framework called Shallow Alignment to improve neural decoding for brain-computer interfaces. This method addresses a granularity mismatch by aligning neural signals with intermediate rep…
-
WaveVerif uses sound to verify robotic workflows with 80% accuracy
Researchers have developed a novel framework called WaveVerif that utilizes acoustic side-channel analysis (ASCA) to verify robotic workflows. This system analyzes the sounds emitted by robots during movement to determi…
-
New AI methods optimize 3D printing quality under uncertainty · 3 sources tracked
Researchers have developed new methodologies for optimizing the fused filament fabrication (FFF) process, focusing on improving part quality under uncertainty. One approach uses Bayesian neural networks to predict geome…
-
New Perspective-Invariant Attack Enhances Adversarial Example Transferability
Researchers have developed a new method called Perspective-Invariant Attack (PIA) to enhance the transferability of adversarial examples in deep neural networks. Unlike previous methods that used limited local transform…