Deep Learning

By Xingbang Hao, Guigang Zhang, Shang Ma. 2016

78,707 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Large Language Models, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Large Language Models

In Signal Processing

Citations in the corpus

Massively Multilingual Neural Machine Translation in the Wild: Findings and Challenges (2019, background): “As with any machine learning problem, the quality and the amount of the data has significant impact on the systems that can be developed (Goodfellow et al., 2016).”
Measuring the Effects of Data Parallelism on Neural Network Training (2018, background): “The dominant algorithms for training neural networks are based on mini-batch stochastic gradient descent (SGD, Robbins and Monro, 1951; Kiefer et al., 1952; Rumelhart et al., 1986; Bottou and Bousquet, 2008; LeCun et al., 2015).”
Measuring and Improving BERT’s Mathematical Abilities by Predicting the Order of Reasoning. (2021, background): “On the other hand, deep learning (LeCun et al., 2015) aims to develop artificial general intelligence that scales better to various problems.”
Overcoming catastrophic forgetting in neural networks (2016, background): “…objectives of task B. Whilst recent advances in machine learning and in particular deep neural networks have resulted in impressive gains in performance across a variety of domains (e.g. [Krizhevsky et al., 2012, LeCun et al., 2015]), little progress has been made in achieving continual learning.”
Data-Efficient Image Recognition with Contrastive Predictive Coding (2019, background): “taset [52]. An alternative hypothesis has therefore proposed that intelligent systems need not be structured a priori, but can instead learn about the structure of the world in an unsupervised manner [6, 26, 36]. Choosing an appropriate training objective is an open problem, but a promising guiding ”
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks (2016, background): “Recent successes in deep learning have shown that neural networks trained by ﬁrst-order gradient based optimization are capable of achieving amazing results in diverse domains like computer vision, speech recognition, and language modelling [5].”
ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness (2018, background): “This notion also appears in other explanations, such as in LeCun et al. (2015): Intermediate CNN layers recognise “parts of familiar objects, and subsequent layers [...] detect objects as combinations of these parts” (p. 436).”
A Study and Comparison of Human and Deep Learning Recognition Performance under Visual Distortions (2017, background): “A more detailed introduction to deep neural networks can be found in [17].”
Training Deep Nets with Sublinear Memory Cost (2016, background): “We have recently witnessed the success of deep neural networks in many domains [8], such as computer vision, speech recognition, natural language processing and reinforcement learning.”
The Cityscapes Dataset for Semantic Urban Scene Understanding (2016, background): “The resurrection of deep learning [34] has had a major impact on the current state-of-the-art in machine learning and computer vision.”
Improved Techniques for Training GANs (2016, background): “For example, when one player minimizes xy with respect to x and another player minimizes − xy with respect to y , gradient descent enters a stable orbit, rather than converging to x = y = 0 , the desired equilibrium point [15].”
Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering (2016, background): “Precisely, CNNs extract the local stationarity property of the input data or signals by revealing local features that are shared across the data domain.”