Bidirectional LSTM Networks for Improved Phoneme Classification and Recognition

By Alex Graves, Santiago Fernández, J. Urgen Schmidhuber. 2005

969 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Model Architectures, Speech & Audio

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 Model Architectures

In Speech & Audio

Citations in the corpus

Ego4D: Around the World in 3,000 Hours of Egocentric Video (2021, background): “[91] Alex Graves, Santiago Fernández, and Jürgen Schmidhuber.”
Temporal Convolutional Networks for Action Segmentation and Detection (2016, methodology): “To make the baselines more competitive, we apply Bidirectional LSTM (Bi-LSTM) [9] to 50 Salads and GTEA.”
Robust Imitation of Diverse Behaviors (2017, methodology): “We first introduce a variational autoencoder (VAE) [15, 26] for supervised imitation, consisting of a bi-directional LSTM [13, 31, 9] encoder mapping demonstration sequences to embedding vectors, and two decoders.”
An Analysis of Incorporating an External Language Model into a Sequence-to-Sequence Model (2017, methodology): “For grapheme experiments, our baseline model is a LAS model with 3 convolutional layers and a convolutional LSTM layer, followed by 3 bidirectional [20] LSTM layers.”
Dynamic Entity Representation with Max-pooling Improves Machine Reading (2016, methodology): “The query vector3 u(q) is constructed similarly as dynamic entity representations, using bidirectional LSTMs4 to encode the query and then encoding the output vectors.”
Pairwise Word Interaction Modeling with Deep Neural Networks for Semantic Similarity Measurement (2016, methodology): “Bidirectional Long Short-Term Memory Networks (Bi-LSTMs) (Graves et al., 2005; Graves et al., 2006) are used for context modeling of input sentences, which serves as the basis for all following components (Sec.”
Adversarial Reprogramming of Text Classification Neural Networks (2018, methodology): “ory (LSTM) network (Hochreiter and Schmidhuber, 1997), bidirectional LSTM network (Graves et al., 2005) and Convolutional neural network (Kim, 2014) models.”
Deep Learning for Entity Matching: A Design Space Exploration (2018, background): “There are many variations of RNN models [46], including long short-term memory (LSTM) networks [33], gated recurrent unit (GRU) [8] networks, and bidirectional networks [29, 49].”
An Application of Recurrent Neural Networks to Discriminative Keyword Spotting (2007, methodology): “The network architecture selected for the keyword spotting task is the bi-directional long short-term memory recurrent neural network (BLSTM), which has shown good performance in a series of speech tasks [12, 13].”
A Review of Deep Learning Models for Time Series Prediction (2019, methodology): “Many conventional architectures utilize only the previous and current timestep, while the bidirectional LSTM utilizes the future timesteps, and was used in sequenceto-sequence speech recognition [100].”
Protein remote homology detection based on bidirectional long short-term memory (2017, methodology): “Here, we are to propose a computational predictor for protein remote homology detection based on Bidirectional Long Short-Term Memory [45, 46, 49], called ProDec-BLSTM , to address the aforementioned disadvantages of the existing methods in this field.”
3-D Convolutional Recurrent Neural Networks With Attention Model for Speech Emotion Recognition (2018, methodology): “bidirectional RNN with long short-term memory cells [19] for temporal summarization, and each direction contains 128 cells; then, we can obtain a sequence of 256-dimensional high-level”
DeClarE: Debunking Fake News and False Claims using Evidence-Aware Deep Learning (2018, methodology): “To create a representation of an article, which may capture task-specific features such as whether it contains objective language, we use a bidirectional Long Short-Term Memory (LSTM) network as proposed by Graves et al. (2005). A basic LSTM cell consists of various gates to control the flow of info”
From Characters to Words to in Between: Do We Capture Morphology? (2017, methodology): “A bi-LSTM (Graves et al., 2005) combines the ﬁnal state of an LSTM over the input sequence with one over the reversed input sequence.”
Semantic Sentence Matching with Densely-connected Recurrent and Co-attentive Information (2018, methodology): “We employ bidirectional LSTM (biLSTM) [11] for preserving sequential information of P and Q.”
Benchmarking of LSTM Networks (2015, background): “LSTM networks [1, 2, 3] have become very popular for many sequence classiﬁ-cation tasks.”
A Multi-stream Bi-directional Recurrent Neural Network for Fine-Grained Action Detection (2016, background): “These are input to bi-directional long short-term memory (LSTM) [8, 6] units to analyze long-term temporal dynamics.”
Deep Air Quality Forecasting Using Hybrid Deep Learning Framework (2018, positive): “ing long -term dependencies. We use a bi-directional LSTM to process time -series information in two directions with two separate hidden layers and then feed this information to the same output layer [29][30] so that it can access both past and future contexts for each point in the time series. 3.2 ”
Learning What’s Easy: Fully Differentiable Neural Easy-First Taggers (2017, methodology): “Part of the success comes from clever architectures such as (bidirectional) long-short term memories (BILSTMs; Hochreiter and Schmidhuber (1997); Graves et al. (2005)) and attention mechanisms (Bahdanau et al., 2015), which are able to select the pieces of context relevant for prediction.”