Model compression and hardware acceleration for neural networks: A comprehensive survey
Domain-specific hardware is becoming a promising topic in the backdrop of improvement
slow down for general-purpose processors due to the foreseeable end of Moore's Law …
slow down for general-purpose processors due to the foreseeable end of Moore's Law …
A survey on deep learning for multimodal data fusion
With the wide deployments of heterogeneous networks, huge amounts of data with
characteristics of high volume, high variety, high velocity, and high veracity are generated …
characteristics of high volume, high variety, high velocity, and high veracity are generated …
Mamba: Linear-time sequence modeling with selective state spaces
Foundation models, now powering most of the exciting applications in deep learning, are
almost universally based on the Transformer architecture and its core attention module …
almost universally based on the Transformer architecture and its core attention module …
Rwkv: Reinventing rnns for the transformer era
Transformers have revolutionized almost all natural language processing (NLP) tasks but
suffer from memory and computational complexity that scales quadratically with sequence …
suffer from memory and computational complexity that scales quadratically with sequence …
[HTML][HTML] Artificial intelligence-enabled detection and assessment of Parkinson's disease using nocturnal breathing signals
There are currently no effective biomarkers for diagnosing Parkinson's disease (PD) or
tracking its progression. Here, we developed an artificial intelligence (AI) model to detect PD …
tracking its progression. Here, we developed an artificial intelligence (AI) model to detect PD …
Simplified state space layers for sequence modeling
Models using structured state space sequence (S4) layers have achieved state-of-the-art
performance on long-range sequence modeling tasks. An S4 layer combines linear state …
performance on long-range sequence modeling tasks. An S4 layer combines linear state …
Combining recurrent, convolutional, and continuous-time models with linear state space layers
Recurrent neural networks (RNNs), temporal convolutions, and neural differential equations
(NDEs) are popular families of deep learning models for time-series data, each with unique …
(NDEs) are popular families of deep learning models for time-series data, each with unique …
Underspecification presents challenges for credibility in modern machine learning
Machine learning (ML) systems often exhibit unexpectedly poor behavior when they are
deployed in real-world domains. We identify underspecification in ML pipelines as a key …
deployed in real-world domains. We identify underspecification in ML pipelines as a key …
Hierarchically gated recurrent neural network for sequence modeling
Transformers have surpassed RNNs in popularity due to their superior abilities in parallel
training and long-term dependency modeling. Recently, there has been a renewed interest …
training and long-term dependency modeling. Recently, there has been a renewed interest …
Delving into deep imbalanced regression
Real-world data often exhibit imbalanced distributions, where certain target values have
significantly fewer observations. Existing techniques for dealing with imbalanced data focus …
significantly fewer observations. Existing techniques for dealing with imbalanced data focus …