Several variants of the long short-term memory (LSTM) architecture for recurrent neural networks have been proposed since its inception in 1995. In recent years, these networks have become the state-of-the-art models for a variety of machine learning problems. This has led to a renewed interest in understanding the role and utility of various computational components of typical LSTM variants. In this paper, we present the first large-scale analysis of eight LSTM variants on three representative tasks: speech recognition, handwriting recognition, and polyphonic music modeling. The hyperparameters of all LSTM variants for each task were optimized separately using random search, and their importance was assessed using the powerful functional ANalysis Of VAriance framework. In total, we summarize the results of 5400 experimental runs ( $\approx 15$ years of CPU time), which makes our study the largest of its kind on LSTM networks. Our results show that none of the variants can improve upon the standard LSTM architecture significantly, and demonstrate the forget gate and the output activation function to be its most critical components. We further observe that the studied hyperparameters are virtually independent and derive guidelines for their efficient adjustment.

LSTM: A Search Space Odyssey

This paper introduces WaveNet, a deep neural network for generating raw audio
waveforms. The model is fully probabilistic and autoregressive, with the
predictive distribution for each audio sample conditioned on all previous ones;
nonetheless we show that it can be efficiently trained on data with tens of
thousands of samples per second of audio. When applied to text-to-speech, it
yields state-of-the-art performance, with human listeners rating it as
significantly more natural sounding than the best parametric and concatenative
systems for both English and Mandarin. A single WaveNet can capture the
characteristics of many different speakers with equal fidelity, and can switch
between them by conditioning on the speaker identity. When trained to model
music, we find that it generates novel and often highly realistic musical
fragments. We also show that it can be employed as a discriminative model,
returning promising results for phoneme recognition.

WaveNet: A Generative Model for Raw Audio

In this paper, we present bidirectional Long Short Term Memory (LSTM) networks, and a modified, full gradient version of the LSTM learning algorithm. We evaluate Bidirectional LSTM (BLSTM) and several other network architectures on the benchmark task of framewise phoneme classification, using the TIMIT database. Our main findings are that bidirectional networks outperform unidirectional ones, and Long Short Term Memory (LSTM) is much faster and also more accurate than both standard Recurrent Neural Nets (RNNs) and time-windowed Multilayer Perceptrons (MLPs). Our results support the view that contextual information is crucial to speech processing, and suggest that BLSTM is an effective architecture with which to exploit it'.

Framewise phoneme classification with bidirectional LSTM and other neural network architectures

Simultaneous tracking of multiple persons in real-world environments is an active research field and several approaches have been proposed, based on a variety of features and algorithms. Recently, there has been a growing interest in organizing systematic evaluations to compare the various techniques. Unfortunately, the lack of common metrics for measuring the performance of multiple object trackers still makes it hard to compare their results. In this work, we introduce two intuitive and general metrics to allow for objective comparison of tracker characteristics, focusing on their precision in estimating object locations, their accuracy in recognizing object configurations and their ability to consistently label objects over time. These metrics have been extensively used in two large-scale international evaluations, the 2006 and 2007 CLEAR evaluations, to measure and compare the performance of multiple object trackers for a wide variety of tracking tasks. Selected performance results are presented and the advantages and drawbacks of the presented metrics are discussed based on the experience gained during the evaluations.

/pdf/evaluating-multiple-object-tracking-performance-the-clear-tm8230vs80.pdf

Evaluating multiple object tracking performance: the CLEAR MOT metrics

2005 Special Issue: Framewise phoneme classification with bidirectional LSTM and other neural network architectures

Darpa Timit Acoustic-Phonetic Continuous Speech Corpus CD-ROM {TIMIT} | NIST

The TIMIT corpus of read speech is designed to provide speech data for acoustic-phonetic studies and for the development and evaluation of automatic speech recognition systems. TIMIT contains broadband recordings of 630 speakers of eight major dialects of American English, each reading ten phonetically rich sentences. The TIMIT corpus includes time-aligned orthographic, phonetic and word transcriptions as well as a 16-bit, 16kHz speech waveform file for each utterance. Corpus design was a joint effort among the Massachusetts Institute of Technology (MIT), SRI International (SRI) and Texas Instruments, Inc. (TI). The speech was recorded at TI, transcribed at MIT and verified and prepared for CD-ROM production by the National Institute of Standards and Technology (NIST). The TIMIT corpus transcriptions have been hand verified. Test and training subsets, balanced for phonetic and dialectal coverage, are specified. Tabular computer-searchable information is included as well as written documentation.

TIMIT Acoustic-Phonetic Continuous Speech Corpus

DARPA TIMIT:: acoustic-phonetic continuous speech corpus CD-ROM, NIST speech disc 1-1.1

This paper describes work within the NIST Text REtrieval Conference (TREC) over the last three years in designing and implementing evaluations of Spoken Document Retrieval (SDR) technology within a broadcast news domain. SDR involves the search and retrieval of excerpts from spoken audio recordings using a combination of automatic speech recognition and information retrieval technologies. The TREC SDR Track has provided an infrastructure for the development and evaluation of SDR technology and a common forum for the exchange of knowledge between the speech recognition and information retrieval research communities. The SDR Track can be declared a success in that it has provided objective, demonstrable proof that this technology can be successfully applied to realistic audio collections using a combination of existing technologies and that it can be objectively evaluated. The design and implementation of each of the SDR evaluations are presented and the results are summarized. Plans for the 2000 TREC SDR Track are presented and thoughts about how the track might evolve are discussed.

The TREC spoken document retrieval track: a success story

paper presents the pilot evaluation of Spoken Term Detection technologies, held during the latter part of 2006. Spoken Term Detection systems rapidly detect the presence of a term , which is a sequence of words consecutively spoken, in a large audio corpus of heterogeneous speech material. The paper describes the evaluation task posed to Spoken Term Detection systems, the evaluation methodologies, the Arabic, English and Mandarin evaluation corpora, and the results of the evaluation. Ten participants submitted systems for the evaluation.

John S. Garofolo

Papers

Darpa Timit Acoustic-Phonetic Continuous Speech Corpus CD-ROM {TIMIT} | NIST

TIMIT Acoustic-Phonetic Continuous Speech Corpus

DARPA TIMIT:: acoustic-phonetic continuous speech corpus CD-ROM, NIST speech disc 1-1.1

The TREC spoken document retrieval track: a success story

Results of the 2006 Spoken Term Detection Evaluation