Attention augmented convolutional networks
Convolutional networks have enjoyed much success in many computer vision applications.
The convolution operation however has a significant weakness in that it only operates on a …
The convolution operation however has a significant weakness in that it only operates on a …
Re-parameterizing your optimizers rather than architectures
The well-designed structures in neural networks reflect the prior knowledge incorporated
into the models. However, though different models have various priors, we are used to …
into the models. However, though different models have various priors, we are used to …
Scalable neural architecture search for 3d medical image segmentation
In this paper, a neural architecture search (NAS) framework is proposed for 3D medical
image segmentation, to automatically optimize a neural architecture from a large design …
image segmentation, to automatically optimize a neural architecture from a large design …
Evolutionary optimization of hyperparameters in deep learning models
Recently, deep learning is one of the most popular techniques in artificial intelligence.
However, to construct a deep learning model, various components must be set up, including …
However, to construct a deep learning model, various components must be set up, including …
Sedona: Search for decoupled neural networks toward greedy block-wise learning
Backward locking and update locking are well-known sources of inefficiency in
backpropagation that prevent from concurrently updating layers. Several works have …
backpropagation that prevent from concurrently updating layers. Several works have …
Software and application patterns for explanation methods
M Alber - Explainable AI: interpreting, explaining and visualizing …, 2019 - Springer
Deep neural networks successfully pervaded many applications domains and are
increasingly used in critical decision processes. Understanding their workings is desirable …
increasingly used in critical decision processes. Understanding their workings is desirable …
Optimizing neural networks through activation function discovery and automatic weight initialization
G Bingham - arXiv preprint arXiv:2304.03374, 2023 - arxiv.org
Automated machine learning (AutoML) methods improve upon existing models by
optimizing various aspects of their design. While present methods focus on hyperparameters …
optimizing various aspects of their design. While present methods focus on hyperparameters …
A two-step rule for backpropagation
A Boughammoura - International Journal of Informatics and Applied …, 2023 - dergipark.org.tr
We present a simplified computational rule for the back-propagation formulas for artificial
neural networks. In this work, we provide a generic two-step rule for the back-propagation …
neural networks. In this work, we provide a generic two-step rule for the back-propagation …
How to iNNvestigate neural networks' predictions!
In recent years, deep neural networks have revolutionized many application domains of
machine learning and are key components of many critical decision or predictive processes …
machine learning and are key components of many critical decision or predictive processes …
Equipment identification through image recognition
D Saidnassimov - 2022 - aaltodoc.aalto.fi
Object detection is a rapidly-evolving field with applications varying from medicine to self-
driving vehicles. As the performance of the deep learning algorithms grow exponentially …
driving vehicles. As the performance of the deep learning algorithms grow exponentially …