TDNN-LSTM model packaging scripts for nnet3 decoder

.gitignore

@@ -1,6 +1,7 @@
 # Compiled extensionless executable files in /src/*/
 # This stanza must precede wildcard patterns below!
 /src/*/*
+!/src/lm/test_data/
 !/src/*/?*.*
 !/src/doc/*
 !/src/*/Makefile

COPYING

@@ -57,30 +57,72 @@ License v 2.0 are set forth below.
 
 Individual Contributors (in alphabetical order)
 
-      Mohit Agarwal      
+      Mohit Agarwal
+      Tanel Alumae
       Gilles Boulianne
       Lukas Burget
+      Dogan Can
+      Guoguo Chen
+      Gaofeng Cheng
       Cisco Corporation
-      Ondrej Glembek
+      Pavel Denisov
+      Ilya Edrenkin
+      Ewald Enzinger
+      Joachim Fainberg
+      Daniel Galvez
+      Pegah Ghahremani
       Arnab Ghoshal
+      Ondrej Glembek
       Go Vivace Inc.
+      Allen Guo
+      Hossein Hadian
+      Lv Hang
       Mirko Hannemann
+      Hendy Irawan
       Navdeep Jaitly
       Johns Hopkins University
+      Shiyin Kang
+      Kirill Katsnelson
+      Tom Ko
+      Danijel Korzinek
+      Gaurav Kumar
+      Ke Li
+      Matthew Maciejewski
+      Vimal Manohar
       Yajie Miao
       Microsoft Corporation
       Petr Motlicek
+      Xingyu Na
+      Vincent Nguyen
+      Lucas Ondel
       Vassil Panayotov
+      Vijayaditya Peddinti
+      Phonexia s.r.o.
+      Ondrej Platek
+      Daniel Povey
+      Yanmin Qian
       Ariya Rastrow
       Saarland University
-      Petr Schwarz      
-      Georg Stemmer
+      Omid Sadjadi
+      Petr Schwarz
+      Yiwen Shao
+      Nickolay V. Shmyrev
       Jan Silovsky
-      Phonexia s.r.o.
-      Yanmin Qian
-      Lucas Ondel
+      Eduardo Silva
+      Peter Smit
+      David Snyder
+      Alexander Solovets
+      Georg Stemmer
+      Pawel Swietojanski
+      Jan "Yenda" Trmal
+      Albert Vernon
       Karel Vesely
+      Yiming Wang
+      Shinji Watanabe
+      Minhua Wu
       Haihua Xu
+      Hainan Xu
+      Xiaohui Zhang
 
 Other Source Material
 

egs/aishell/README.txt

@@ -5,5 +5,9 @@ Aishell is an open Chinese Mandarin speech database published by Beijing Shell S
 The database can be downloaded from openslr:
 http://www.openslr.org/33/
 
+This folder contains two subfolders:
+s5: a speech recognition recipe
+v1: a speaker recognition recipe
+
 For more details, please visit:
 http://www.aishelltech.com/kysjcp

egs/aishell/v1/README.txt

@@ -0,0 +1,8 @@
+The database can be downloaded from openslr:
+http://www.openslr.org/33/
+
+For more details, please visit:
+http://www.aishelltech.com/kysjcp
+
+We use the training set to train model,
+and split the test set into enroll and eval.

egs/aishell/v1/cmd.sh

@@ -0,0 +1,15 @@
+# you can change cmd.sh depending on what type of queue you are using.
+# If you have no queueing system and want to run on a local machine, you
+# can change all instances 'queue.pl' to run.pl (but be careful and run
+# commands one by one: most recipes will exhaust the memory on your
+# machine).  queue.pl works with GridEngine (qsub).  slurm.pl works
+# with slurm.  Different queues are configured differently, with different
+# queue names and different ways of specifying things like memory;
+# to account for these differences you can create and edit the file
+# conf/queue.conf to match your queue's configuration.  Search for
+# conf/queue.conf in http://kaldi-asr.org/doc/queue.html for more information,
+# or search for the string 'default_config' in utils/queue.pl or utils/slurm.pl.
+
+export train_cmd="queue.pl --mem 4G"
+
+

egs/aishell/v1/conf/mfcc.conf

@@ -0,0 +1,3 @@
+--sample-frequency=16000
+--num-mel-bins=40 #higher than the default which is 23
+--num-ceps=20 # higher than the default which is 12.

egs/aishell/v1/conf/vad.conf

@@ -0,0 +1,2 @@
+--vad-energy-threshold=5.5
+--vad-energy-mean-scale=0.5

egs/aishell/v1/local/aishell_data_prep.sh

@@ -0,0 +1,63 @@
+#!/bin/bash
+
+# Copyright 2017 Xingyu Na
+# Apache 2.0
+
+. ./path.sh || exit 1;
+
+if [ $# != 2 ]; then
+  echo "Usage: $0 <audio-path> <text-path>"
+  echo " $0 /export/a05/xna/data/data_aishell/wav /export/a05/xna/data/data_aishell/transcript"
+  exit 1;
+fi
+
+aishell_audio_dir=$1
+aishell_text_dir=$2
+
+train_dir=data/local/train
+dev_dir=data/local/dev
+test_dir=data/local/test
+
+mkdir -p $train_dir
+mkdir -p $dev_dir
+mkdir -p $test_dir
+
+# data directory check
+if [ ! -d $aishell_audio_dir ] || [ ! -d $aishell_text_dir ]; then
+  echo "Error: $0 requires two directory arguments"
+  exit 1;
+fi
+
+# find wav audio file for train, dev and test resp.
+find $aishell_audio_dir -iname "*.wav" | grep -i "wav/train" > $train_dir/wav.flist || exit 1;
+find $aishell_audio_dir -iname "*.wav" | grep -i "wav/dev" > $dev_dir/wav.flist || exit 1;
+find $aishell_audio_dir -iname "*.wav" | grep -i "wav/test" > $test_dir/wav.flist || exit 1;
+
+n=`cat $train_dir/wav.flist $dev_dir/wav.flist $test_dir/wav.flist | wc -l`
+[ $n -ne 141925 ] && \
+  echo Warning: expected 141925 data data files, found $n
+
+# Transcriptions preparation
+for dir in $train_dir $test_dir; do
+  echo Preparing $dir transcriptions
+  sed -e 's/\.wav//' $dir/wav.flist | awk -F '/' '{print $NF}' |\
+    sort > $dir/utt.list
+  sed -e 's/\.wav//' $dir/wav.flist | awk -F '/' '{i=NF-1;printf("%s %s\n",$NF,$i)}' |\
+    sort > $dir/utt2spk_all
+  paste -d' ' $dir/utt.list $dir/wav.flist > $dir/wav.scp_all
+  utils/filter_scp.pl -f 1 $dir/utt.list $aishell_text_dir/*.txt > $dir/transcripts.txt
+  awk '{print $1}' $dir/transcripts.txt > $dir/utt.list
+  utils/filter_scp.pl -f 1 $dir/utt.list $dir/utt2spk_all | sort -u > $dir/utt2spk
+  utils/filter_scp.pl -f 1 $dir/utt.list $dir/wav.scp_all | sort -u > $dir/wav.scp
+  sort -u $dir/transcripts.txt > $dir/text
+  utils/utt2spk_to_spk2utt.pl $dir/utt2spk > $dir/spk2utt
+done
+
+mkdir -p data/train data/test
+for f in spk2utt utt2spk wav.scp text; do
+  cp $train_dir/$f data/train/$f || exit 1;
+  cp $test_dir/$f data/test/$f || exit 1;
+done
+
+echo "$0: AISHELL data preparation succeeded"
+exit 0;

egs/aishell/v1/local/download_and_untar.sh

@@ -0,0 +1,105 @@
+#!/bin/bash
+
+# Copyright   2014  Johns Hopkins University (author: Daniel Povey)
+#             2017  Xingyu Na
+# Apache 2.0
+
+remove_archive=false
+
+if [ "$1" == --remove-archive ]; then
+  remove_archive=true
+  shift
+fi
+
+if [ $# -ne 3 ]; then
+  echo "Usage: $0 [--remove-archive] <data-base> <url-base> <corpus-part>"
+  echo "e.g.: $0 /export/a05/xna/data www.openslr.org/resources/33 data_aishell"
+  echo "With --remove-archive it will remove the archive after successfully un-tarring it."
+  echo "<corpus-part> can be one of: data_aishell, resource."
+fi
+
+data=$1
+url=$2
+part=$3
+
+if [ ! -d "$data" ]; then
+  echo "$0: no such directory $data"
+  exit 1;
+fi
+
+part_ok=false
+list="data_aishell resource"
+for x in $list; do
+  if [ "$part" == $x ]; then part_ok=true; fi
+done
+if ! $part_ok; then
+  echo "$0: expected <corpus-part> to be one of $list, but got '$part'"
+  exit 1;
+fi
+
+if [ -z "$url" ]; then
+  echo "$0: empty URL base."
+  exit 1;
+fi
+
+if [ -f $data/$part/.complete ]; then
+  echo "$0: data part $part was already successfully extracted, nothing to do."
+  exit 0;
+fi
+
+# sizes of the archive files in bytes.
+sizes="15582913665 1246920"
+
+if [ -f $data/$part.tgz ]; then
+  size=$(/bin/ls -l $data/$part.tgz | awk '{print $5}')
+  size_ok=false
+  for s in $sizes; do if [ $s == $size ]; then size_ok=true; fi; done
+  if ! $size_ok; then
+    echo "$0: removing existing file $data/$part.tgz because its size in bytes $size"
+    echo "does not equal the size of one of the archives."
+    rm $data/$part.gz
+  else
+    echo "$data/$part.tgz exists and appears to be complete."
+  fi
+fi
+
+if [ ! -f $data/$part.tgz ]; then
+  if ! which wget >/dev/null; then
+    echo "$0: wget is not installed."
+    exit 1;
+  fi
+  full_url=$url/$part.tgz
+  echo "$0: downloading data from $full_url.  This may take some time, please be patient."
+
+  cd $data
+  if ! wget --no-check-certificate $full_url; then
+    echo "$0: error executing wget $full_url"
+    exit 1;
+  fi
+fi
+
+cd $data
+
+if ! tar -xvzf $part.tgz; then
+  echo "$0: error un-tarring archive $data/$part.tgz"
+  exit 1;
+fi
+
+touch $data/$part/.complete
+
+if [ $part == "data_aishell" ]; then
+  cd $data/$part/wav
+  for wav in ./*.tar.gz; do
+    echo "Extracting wav from $wav"
+    tar -zxf $wav && rm $wav
+  done
+fi
+
+echo "$0: Successfully downloaded and un-tarred $data/$part.tgz"
+
+if $remove_archive; then
+  echo "$0: removing $data/$part.tgz file since --remove-archive option was supplied."
+  rm $data/$part.tgz
+fi
+
+exit 0;

egs/aishell/v1/local/produce_trials.py

@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+# Copyright 2017 Bengu Wu
+# Apache 2.0.
+
+# This script generate trials file.
+# Trial file is formatted as:
+# uttid spkid target|nontarget
+
+# If uttid belong to spkid, it is marked 'target',
+# otherwise is 'nontarget'.
+# input: eval set uttspk file
+# output: trial file
+
+import sys
+
+fnutt = sys.argv[1]
+ftrial = open(sys.argv[2], 'w')
+
+dictutt = {}
+for line in open(fnutt):
+  utt2spk = line.rstrip('\r\t\n ')
+  spk = utt2spk.split(' ')[1]
+  if spk not in dictutt:
+    dictutt[spk] = spk
+
+for line in open(fnutt):
+  utt2spk = line.rstrip('\r\t\n ')
+  utt, spk = utt2spk.split(' ')
+  for target in dictutt:
+    if target == spk:
+      trial = utt + ' ' + target + ' target'
+    else:
+      trial = utt + ' ' + target + ' nontarget'
+    ftrial.write(trial + '\n')
+ftrial.close()

egs/aishell/v1/local/split_data_enroll_eval.py

@@ -0,0 +1,37 @@
+#!/usr/bin/env python3
+
+# Copyright 2017 Bengu Wu
+# Apache 2.0.
+
+# This script splits the test set utt2spk into enroll set and eval set
+# For each speaker, 3 utterances are randomly selected as enroll samples,
+# and the others are used as eval samples for evaluation
+# input: test utt2spk
+# output: enroll utt2spk, eval utt2spk
+
+import sys,random
+
+dictutt = {}
+
+for line in open(sys.argv[1]):
+  line = line.rstrip('\r\t\n ')
+  utt, spk = line.split(' ')
+  if spk not in dictutt:
+    dictutt[spk] = []
+  dictutt[spk].append(utt)
+
+fenroll = open(sys.argv[2], 'w')
+feval = open(sys.argv[3], 'w')
+
+for key in dictutt:
+  utts = dictutt[key]
+  random.shuffle(utts)
+  for i in range(0, len(utts)):
+    line = utts[i] + ' ' + key
+    if(i < 3):
+      fenroll.write(line + '\n')
+    else:
+      feval.write(line + '\n')
+
+fenroll.close()
+feval.close()

egs/aishell/v1/path.sh

@@ -0,0 +1,5 @@
+export KALDI_ROOT=`pwd`/../../..
+export PATH=$PWD/utils/:$KALDI_ROOT/tools/openfst/bin:$KALDI_ROOT/tools/sph2pipe_v2.5:$PWD:$PATH
+[ ! -f $KALDI_ROOT/tools/config/common_path.sh ] && echo >&2 "The standard file $KALDI_ROOT/tools/config/common_path.sh is not present -> Exit!" && exit 1
+. $KALDI_ROOT/tools/config/common_path.sh
+export LC_ALL=C