Unconstrained Online Handwriting Recognition with Recurrent Neural Networks

By Alex Graves, Marcus Liwicki. 2007

Subfields: Computer Vision and Pattern Recognition, Model Architectures

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Model Architectures

Citations in the corpus

Speech recognition with deep recurrent neural networks (2013, background): “The combination of Long Short-term Memory [11], an RNN architecture with an improved memory, with end-to-end training has proved especially effective, yielding state-of-the-art results in cursive handwriting recognition [12, 13].”
Sequence Transduction with Recurrent Neural Networks (2012, background): “Furthermore, these results demonstrate the use of long-range memory to perform such actions as closing parentheses after many intervening characters (Sutskever et al., 2011), or using delayed strokes to identify handwritten characters from pen trajectories (Graves et al., 2008).”
Learning Stochastic Recurrent Networks (2014, background): “While only bearing marginal existence in the 1990’s, recent successes in real world applications (Graves, 2013; Graves et al., 2013; Sutskever et al., 2014; Graves et al., 2008; Cho et al., 2014) have resurged interest.”
Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks (2018, methodology): “The LSTM has been successfully applied in sequence modeling [32], speech recognition [28–30], hand-writing recognition [31] and language translation [33].”
Supervised Sequence Labelling with Recurrent Neural Networks (2012, unknown): “, 2006) and handwriting recognition (Liwicki et al., 2007; Graves et al., 2008).”
Learning Representations from EEG with Deep Recurrent-Convolutional Neural Networks (2015, background): “On the other hand, recurrent neural networks have delivered state-of-the-art performance in many applications involving dynamics in temporal sequences, such as, for example, handwriting and speech recognition (Graves et al., 2013; 2008).”
Offline Handwriting Recognition with Multidimensional Recurrent Neural Networks (2008, methodology): “Furthermore, recent results suggest that recurrent neural networks (RNNs) may be preferable to HMMs for sequence labelling tasks such as speech [5] and online handwriting recognition [6].”
From speech to letters - using a novel neural network architecture for grapheme based ASR (2009, methodology): “Using the CTC token passing algorithm [3], the decoding procedure can be extended to incorporate dictionaries and language models.”
Offline Handwriting Recognition with Multidimensional Recurrent Neural Networks (2008, unknown): “Following our recent success in transcribing raw online handwriting data with recurrent networks [6], we wanted to build an offline recognition system that would work on raw pixels.”
Understanding LSTM - a tutorial into Long Short-Term Memory Recurrent Neural Networks (2019, methodology): “In the following sections we cover the most common in use, namely bidirectional LSTM (BLSTM-CTC) ([34], [27], [31]), Grid LSTM (or N-LSTM) [49] and Gated Recurrent Unit (GRU) ([10], [13]).”
Variational inference of latent state sequences using Recurrent Networks (2014, background): “In the simplest cases, the individual factors of Equation (1) are factorised Bernoulli or Gaussian distributions.”
Semi-supervised Learning with Encoder-Decoder Recurrent Neural Networks: Experiments with Motion Capture Sequences (2015, background): “…very powerful on multiple sequential problems, such as speech recognition (Graves et al., 2013b; Sak et al., 2014; Graves et al., 2013a), handwriting recognition (Graves et al., 2008), text generation (Sutskever et al., 2011; Graves, 2012), or in our case MOCAP action recognition (Du et al., 2015).”
On Fast Dropout and its Applicability to Recurrent Networks (2013, methodology): “These approaches are among the best methods for the modelling of speech and handwriting (Graves et al., 2013, 2008; Graves, 2013).”
A Novel Connectionist System for Unconstrained Handwriting Recognition (2009, methodology): “The network uses the connectionist temporal classification (CTC) output layer [31], [32], first applied to speech recognition.”
Improving Offline Handwritten Text Recognition with Hybrid HMM/ANN Models (2011, unknown): “…HMM/ANN models for automatic speech recognition [13], [14], [15], [16], [17] and for online handwriting recognition [18], [19], [20], [21], [22], [23], . and previous works which used hybrid HMM/ ANN word models with remarkable success, although they were limited to digit recognition or small…”