笔记摘录：2018.09.01---Kaldi构建一个简单的英文数字串识别系统

2023-11-22 大全 28 作者：考证青年

本文主要参考的是，主要讲述的是用自己的录音来构建一个数字串识别系统。

本文将主要分为以下几个部分：

识别英语文字__识别英文句子

这里是英文数字串识别，因此需要一些用英语朗读数字的语音。我录制了 128 个语音文件，分别是两个人朗读，其中每个文件只包含三个数字。这 128 文件中 80 个用于训练， 48 个用于测试。并且训练数据和测试数据都被分成了 8 部分（可以假装成 8 个人），每部分分别 10 个和 6 个。读者可以到我的上下载这些语音数据。训练集和测试集的前五个目录是我(男)朗读，后面三个是女士朗读的。

我的录音是 16 kHz 采样，16 位量化的。如果是 8 kHz 采样，需要在提特征时指明采样频率。即将 steps/.sh 里面 -mfcc-feats 命令加上 ---=8000 的选项。

目录结构如下：

识别英文句子_识别英语文字_

当然您也可以选择自己录音，也欢迎你把你的录音共享。

从数据准备到编写脚本一直都是在为最后做准备，如果都完成了的话，你的目录应该长这样：

识别英语文字__识别英文句子

数据准备

数据的准备包括声学数据和语言数据。首先建立一个目录 data/。

声学数据

声学数据主要包括训练集和测试集的数据，以及一个语料库，在 data/ 目录下新建两个目录：train/ 、test/ 和 local。

声学训练数据准备

这些数据主要是四个文件：、wav.scp、text 和，都保存在 data/train/ 目录下。

这个文件主要描述说话人编号和性别的对应关系，具有这样的形式：，比如我这里训练集是“8”个人，前 5 个男(Male)，后 3 个女()。比如：

wav.scp

这个文件主要是保存了语音编号和语音文件的对应关系，具有这样的形式：。比如：

_识别英文句子_识别英语文字

text

识别英文句子_识别英语文字_

这个文件主要是标注，描述了语音编号和标注之间的对应关系，具有这样的形式：。比如

识别英文句子__识别英语文字

这个文件是联系说话人编号和语音编号，具有这样的形式：。比如：

这里我们建议把说话人的编号放在语音编号的前缀。至此，训练集的声学数据准备好了。

声学测试数据准备

和训练集数据一样，这些数据也主要是四个文件：、wav.scp、text 和，都保存在 data/test/ 目录下。这些目录的形式和之前一样，只是 wav.scp 需要映射到测试集的数据，重新修改语音编号。

语料库

我们还需要一个语料库，其包含了所有训练集数据的标注，请命名为 .txt，并保存在 data/local/ 目录下。比如：

_识别英文句子_识别英语文字

语言数据

语言数据主要包括 .txt、.txt、.txt、.txt，并在 data/local/ 目录下新建文件夹 dict，将这四个文件保存到那里。

.txt

这个文件应该包括你标注里所有出现的词的发音，即音素表达，由于这里只有十个单词，再加上静音，因此不管是词还是音素，数量都比较少。

识别英文句子_识别英语文字_

.txt

这个文件列出了上面出现的所有的非静音音素。

.txt

这里面包含了静音音素。

.txt

_识别英语文字_识别英文句子

只有可选的静音音素。

环境准备

回到项目根目录。首先我们需要定义文件 cmd.sh 和 path.sh，其中前者主要包括运行的形式，而后者主要包括 kaldi 依赖的路径。

cmd.sh：

export train_cmd="run.pl"
export decode_cmd="run.pl"
export mkgraph_cmd="run.pl"

path.sh：

export train_cmd="run.pl"
export decode_cmd="run.pl"
export cuda_cmd="run.pl"
export mkgraph_cmd="run.pl"
Huang-Lus-MacBook-Air:en huanglu$ cat path.sh 
export KALDI_ROOT=`pwd`/../../..
[ -f $KALDI_ROOT/tools/env.sh ] && . $KALDI_ROOT/tools/env.sh
export PATH=$PWD/utils/:$KALDI_ROOT/tools/openfst/bin:$KALDI_ROOT/tools/irstlm/bin/:$PWD:$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
Huang-Lus-MacBoo

此外我们还需要一些 kaldi 已经写好的脚本，比如 steps/.sh 和 utils/.sh，简单起见，我们直接这两个文件夹链接过来。

$ ln -s ../../timit/s5/steps steps
$ ln -s ../../timit/s5/utils utils

然后就是为了打分以及处理提特征时采样率非 16 kHz 时，需要在根目录下建立 local 文件夹，然后从别的那里拷贝来 .sh 和 score.sh。

最后就是建立目录 conf，存放一些配置文件。包括：

.：

first_beam=10.0
beam=13.0
lattice_beam=6.0

mfcc.conf：

--use-energy=false

编写运行脚本

在根目录下建立一个 run.sh 脚本，输入以下内容（有注释，我不解释了）：

#!/bin/bash. ./path.sh || exit 1
. ./cmd.sh || exit 1nj=4
lm_order=1. utils/parse_options.sh || exit 1
[[ $# -ge 1 ]] && { echo "Wrong arguments!"; exit 1; }# Removing previously created data (from last run.sh execution)
rm -rf exp mfcc data/train/spk2utt data/train/cmvn.scp data/train/feats.scp \data/train/split1 data/test/spk2utt data/test/cmvn.scp data/test/feats.scp \data/test/split1 data/local/lang data/lang data/local/tmp \data/local/dict/lexiconp.txtecho
echo "===== PREPARING ACOUSTIC DATA ====="
echo# Making spk2utt files
utils/utt2spk_to_spk2utt.pl data/train/utt2spk > data/train/spk2utt
utils/utt2spk_to_spk2utt.pl data/test/utt2spk > data/test/spk2uttecho
echo "===== FEATURES EXTRACTION ====="
echo# Making feats.scp files
mfccdir=mfcc
# Uncomment and modify arguments in scripts below if you have any problems with data sorting
# utils/validate_data_dir.sh data/train     # script for checking prepared data - here: for data/train directory
# utils/validate_data_dir.sh data/test
# utils/fix_data_dir.sh data/train          # tool for data proper sorting if needed - here: for data/train directory
# utils/fix_data_dir.sh data/teststeps/make_mfcc.sh --nj $nj --cmd "$train_cmd" data/train \exp/make_mfcc/train $mfccdir
steps/make_mfcc.sh --nj $nj --cmd "$train_cmd" data/test \exp/make_mfcc/test $mfccdir# Making cmvn.scp files
steps/compute_cmvn_stats.sh data/train exp/make_mfcc/train $mfccdir
steps/compute_cmvn_stats.sh data/test exp/make_mfcc/test $mfccdirecho
echo "===== PREPARING LANGUAGE DATA ====="
echo
# Preparing language data
utils/prepare_lang.sh data/local/dict "" data/local/lang data/langecho
echo "===== LANGUAGE MODEL CREATION ====="
echo "===== MAKING lm.arpa ====="
echoloc=`which ngram-count`;
if [ -z $loc ]; thenif uname -a | grep 64 >/dev/null; thensdir=$KALDI_ROOT/tools/srilm/bin/i686-m64elsesdir=$KALDI_ROOT/tools/srilm/bin/i686fiif [ -f $sdir/ngram-count ]; thenecho "Using SRILM language modelling tool from $sdir"export PATH=$PATH:$sdirelseecho "SRILM toolkit is probably not installed. Instructions: tools/install_srilm.sh"exit 1fi
filocal=data/local
mkdir $local/tmp
ngram-count -order $lm_order -write-vocab $local/tmp/vocab-full.txt \-wbdiscount -text $local/corpus.txt -lm $local/tmp/lm.arpaecho
echo "===== MAKING G.fst ====="
echolang=data/lang
arpa2fst --disambig-symbol=#0 --read-symbol-table=$lang/words.txt \$local/tmp/lm.arpa $lang/G.fstecho
echo "===== MONO TRAINING ====="
echosteps/train_mono.sh --nj $nj --cmd "$train_cmd" data/train data/lang exp/mono  || exit 1echo
echo "===== MONO DECODING ====="
echoutils/mkgraph.sh --mono data/lang exp/mono exp/mono/graph || exit 1
steps/decode.sh --config conf/decode.config --nj $nj --cmd "$decode_cmd" \exp/mono/graph data/test exp/mono/decode
local/score.sh data/test data/lang exp/mono/decode/echo
echo "===== MONO ALIGNMENT ====="
echosteps/align_si.sh --nj $nj --cmd "$train_cmd" data/train data/lang exp/mono exp/mono_ali || exit 1echo
echo "===== TRI1 (first triphone pass) TRAINING ====="
echosteps/train_deltas.sh --cmd "$train_cmd" 2000 11000 data/train data/lang exp/mono_ali exp/tri1 || exit 1echo
echo "===== TRI1 (first triphone pass) DECODING ====="
echoutils/mkgraph.sh data/lang exp/tri1 exp/tri1/graph || exit 1
steps/decode.sh --config conf/decode.config --nj $nj --cmd "$decode_cmd" \exp/tri1/graph data/test exp/tri1/decode
local/score.sh data/test data/lang exp/tri1/decode/echo
echo "===== run.sh script is finished ====="
echo

运行脚本

最后在运行 run.sh 之前先进行下面的命令：

$ chmod u+x local/*.sh *.sh

运行结果（省去了部分重复的）：

$ ./run.sh ===== PREPARING ACOUSTIC DATA ========== FEATURES EXTRACTION =====steps/make_mfcc.sh --nj 4 --cmd run.pl data/train exp/make_mfcc/train mfcc
utils/validate_data_dir.sh: Successfully validated data-directory data/train
steps/make_mfcc.sh: [info]: no segments file exists: assuming wav.scp indexed by utterance.
Succeeded creating MFCC features for train
steps/make_mfcc.sh --nj 4 --cmd run.pl data/test exp/make_mfcc/test mfcc
utils/validate_data_dir.sh: Successfully validated data-directory data/test
steps/make_mfcc.sh: [info]: no segments file exists: assuming wav.scp indexed by utterance.
Succeeded creating MFCC features for test
steps/compute_cmvn_stats.sh data/train exp/make_mfcc/train mfcc
Succeeded creating CMVN stats for train
steps/compute_cmvn_stats.sh data/test exp/make_mfcc/test mfcc
Succeeded creating CMVN stats for test===== PREPARING LANGUAGE DATA =====utils/prepare_lang.sh data/local/dict  data/local/lang data/lang
Checking data/local/dict/silence_phones.txt ...
--> reading data/local/dict/silence_phones.txt
--> data/local/dict/silence_phones.txt is OKChecking data/local/dict/optional_silence.txt ...
--> reading data/local/dict/optional_silence.txt
--> data/local/dict/optional_silence.txt is OKChecking data/local/dict/nonsilence_phones.txt ...
--> reading data/local/dict/nonsilence_phones.txt
--> data/local/dict/nonsilence_phones.txt is OKChecking disjoint: silence_phones.txt, nonsilence_phones.txt
--> disjoint property is OK.Checking data/local/dict/lexicon.txt
--> reading data/local/dict/lexicon.txt
--> data/local/dict/lexicon.txt is OKChecking data/local/dict/extra_questions.txt ...
--> data/local/dict/extra_questions.txt is empty (this is OK)
--> SUCCESS [validating dictionary directory data/local/dict]**Creating data/local/dict/lexiconp.txt from data/local/dict/lexicon.txt
fstaddselfloops data/lang/phones/wdisambig_phones.int data/lang/phones/wdisambig_words.int 
prepare_lang.sh: validating output directory
utils/validate_lang.pl data/lang
Checking data/lang/phones.txt ...
--> data/lang/phones.txt is OKChecking words.txt: #0 ...
--> data/lang/words.txt is OKChecking disjoint: silence.txt, nonsilence.txt, disambig.txt ...
--> silence.txt and nonsilence.txt are disjoint
--> silence.txt and disambig.txt are disjoint
--> disambig.txt and nonsilence.txt are disjoint
--> disjoint property is OKChecking sumation: silence.txt, nonsilence.txt, disambig.txt ...
--> summation property is OKChecking data/lang/phones/context_indep.{txt, int, csl} ...
--> 10 entry/entries in data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.int corresponds to data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.csl corresponds to data/lang/phones/context_indep.txt
--> data/lang/phones/context_indep.{txt, int, csl} are OKChecking data/lang/phones/nonsilence.{txt, int, csl} ...
--> 80 entry/entries in data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.int corresponds to data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.csl corresponds to data/lang/phones/nonsilence.txt
--> data/lang/phones/nonsilence.{txt, int, csl} are OKChecking data/lang/phones/silence.{txt, int, csl} ...
--> 10 entry/entries in data/lang/phones/silence.txt
--> data/lang/phones/silence.int corresponds to data/lang/phones/silence.txt
--> data/lang/phones/silence.csl corresponds to data/lang/phones/silence.txt
--> data/lang/phones/silence.{txt, int, csl} are OKChecking data/lang/phones/optional_silence.{txt, int, csl} ...
--> 1 entry/entries in data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.int corresponds to data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.csl corresponds to data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.{txt, int, csl} are OKChecking data/lang/phones/disambig.{txt, int, csl} ...
--> 2 entry/entries in data/lang/phones/disambig.txt
--> data/lang/phones/disambig.int corresponds to data/lang/phones/disambig.txt
--> data/lang/phones/disambig.csl corresponds to data/lang/phones/disambig.txt
--> data/lang/phones/disambig.{txt, int, csl} are OKChecking data/lang/phones/roots.{txt, int} ...
--> 22 entry/entries in data/lang/phones/roots.txt
--> data/lang/phones/roots.int corresponds to data/lang/phones/roots.txt
--> data/lang/phones/roots.{txt, int} are OKChecking data/lang/phones/sets.{txt, int} ...
--> 22 entry/entries in data/lang/phones/sets.txt
--> data/lang/phones/sets.int corresponds to data/lang/phones/sets.txt
--> data/lang/phones/sets.{txt, int} are OKChecking data/lang/phones/extra_questions.{txt, int} ...
--> 9 entry/entries in data/lang/phones/extra_questions.txt
--> data/lang/phones/extra_questions.int corresponds to data/lang/phones/extra_questions.txt
--> data/lang/phones/extra_questions.{txt, int} are OKChecking data/lang/phones/word_boundary.{txt, int} ...
--> 90 entry/entries in data/lang/phones/word_boundary.txt
--> data/lang/phones/word_boundary.int corresponds to data/lang/phones/word_boundary.txt
--> data/lang/phones/word_boundary.{txt, int} are OKChecking optional_silence.txt ...
--> reading data/lang/phones/optional_silence.txt
--> data/lang/phones/optional_silence.txt is OKChecking disambiguation symbols: #0 and #1
--> data/lang/phones/disambig.txt has "#0" and "#1"
--> data/lang/phones/disambig.txt is OKChecking topo ...Checking word_boundary.txt: silence.txt, nonsilence.txt, disambig.txt ...
--> data/lang/phones/word_boundary.txt doesn't include disambiguation symbols
--> data/lang/phones/word_boundary.txt is the union of nonsilence.txt and silence.txt
--> data/lang/phones/word_boundary.txt is OKChecking word-level disambiguation symbols...
--> data/lang/phones/wdisambig.txt exists (newer prepare_lang.sh)
Checking word_boundary.int and disambig.int
--> generating a 67 word sequence
--> resulting phone sequence from L.fst corresponds to the word sequence
--> L.fst is OK
--> generating a 16 word sequence
--> resulting phone sequence from L_disambig.fst corresponds to the word sequence
--> L_disambig.fst is OKChecking data/lang/oov.{txt, int} ...
--> 1 entry/entries in data/lang/oov.txt
--> data/lang/oov.int corresponds to data/lang/oov.txt
--> data/lang/oov.{txt, int} are OK--> data/lang/L.fst is olabel sorted
--> data/lang/L_disambig.fst is olabel sorted
--> SUCCESS [validating lang directory data/lang]===== LANGUAGE MODEL CREATION =====
===== MAKING lm.arpa ========== MAKING G.fst =====arpa2fst --disambig-symbol=#0 --read-symbol-table=data/lang/words.txt data/local/tmp/lm.arpa data/lang/G.fst 
LOG (arpa2fst:Read():arpa-file-parser.cc:96) Reading \data\ section.
LOG (arpa2fst:Read():arpa-file-parser.cc:151) Reading \1-grams: section.
LOG (arpa2fst:RemoveRedundantStates():arpa-lm-compiler.cc:355) Reduced num-states from 1 to 1===== MONO TRAINING =====steps/train_mono.sh --nj 4 --cmd run.pl data/train data/lang exp/mono
steps/train_mono.sh: Initializing monophone system.
steps/train_mono.sh: Compiling training graphs
steps/train_mono.sh: Aligning data equally (pass 0)
steps/train_mono.sh: Pass 1
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 2
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 3
···············
steps/train_mono.sh: Pass 38
steps/train_mono.sh: Aligning data
steps/train_mono.sh: Pass 39
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/mono
steps/diagnostic/analyze_alignments.sh: see stats in exp/mono/log/analyze_alignments.log
311 warnings in exp/mono/log/update.*.log
84 warnings in exp/mono/log/align.*.*.log
exp/mono: nj=4 align prob=-79.54 over 0.06h [retry=0.0%, fail=0.0%] states=70 gauss=1007
steps/train_mono.sh: Done training monophone system in exp/mono===== MONO DECODING =====WARNING: the --mono, --left-biphone and --quinphone options are now deprecated and ignored.
tree-info exp/mono/tree 
tree-info exp/mono/tree 
fstdeterminizestar --use-log=true 
fsttablecompose data/lang/L_disambig.fst data/lang/G.fst 
fstminimizeencoded 
fstpushspecial 
fstisstochastic data/lang/tmp/LG.fst 
-0.0421835 -0.0423222
[info]: LG not stochastic.
fstcomposecontext --context-size=1 --central-position=0 --read-disambig-syms=data/lang/phones/disambig.int --write-disambig-syms=data/lang/tmp/disambig_ilabels_1_0.int data/lang/tmp/ilabels_1_0 
fstisstochastic data/lang/tmp/CLG_1_0.fst 
-0.0421835 -0.0423222
[info]: CLG not stochastic.
make-h-transducer --disambig-syms-out=exp/mono/graph/disambig_tid.int --transition-scale=1.0 data/lang/tmp/ilabels_1_0 exp/mono/tree exp/mono/final.mdl 
fsttablecompose exp/mono/graph/Ha.fst data/lang/tmp/CLG_1_0.fst 
fstminimizeencoded 
fstrmepslocal 
fstdeterminizestar --use-log=true 
fstrmsymbols exp/mono/graph/disambig_tid.int 
fstisstochastic exp/mono/graph/HCLGa.fst 
0.000319138 -0.0427566
HCLGa is not stochastic
add-self-loops --self-loop-scale=0.1 --reorder=true exp/mono/final.mdl 
steps/decode.sh --config conf/decode.config --nj 4 --cmd run.pl exp/mono/graph data/test exp/mono/decode
decode.sh: feature type is delta
steps/diagnostic/analyze_lats.sh --cmd run.pl exp/mono/graph exp/mono/decode
steps/diagnostic/analyze_lats.sh: see stats in exp/mono/decode/log/analyze_alignments.log
Overall, lattice depth (10,50,90-percentile)=(1,1,2) and mean=1.2
steps/diagnostic/analyze_lats.sh: see stats in exp/mono/decode/log/analyze_lattice_depth_stats.log
exp/mono/decode/wer_10
%WER 1.39 [ 2 / 144, 1 ins, 0 del, 1 sub ]
%SER 4.17 [ 2 / 48 ]
exp/mono/decode/wer_11
%WER 1.39 [ 2 / 144, 1 ins, 0 del, 1 sub ]
%SER 4.17 [ 2 / 48 ]
exp/mono/decode/wer_12
····················
%SER 4.17 [ 2 / 48 ]
exp/mono/decode//wer_8
%WER 1.39 [ 2 / 144, 1 ins, 0 del, 1 sub ]
%SER 4.17 [ 2 / 48 ]
exp/mono/decode//wer_9
%WER 1.39 [ 2 / 144, 1 ins, 0 del, 1 sub ]
%SER 4.17 [ 2 / 48 ]===== MONO ALIGNMENT =====steps/align_si.sh --nj 4 --cmd run.pl data/train data/lang exp/mono exp/mono_ali
steps/align_si.sh: feature type is delta
steps/align_si.sh: aligning data in data/train using model from exp/mono, putting alignments in exp/mono_ali
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/mono_ali
steps/diagnostic/analyze_alignments.sh: see stats in exp/mono_ali/log/analyze_alignments.log
steps/align_si.sh: done aligning data.===== TRI1 (first triphone pass) TRAINING =====steps/train_deltas.sh --cmd run.pl 2000 11000 data/train data/lang exp/mono_ali exp/tri1
steps/train_deltas.sh: accumulating tree stats
steps/train_deltas.sh: getting questions for tree-building, via clustering
steps/train_deltas.sh: building the tree
WARNING (gmm-init-model:InitAmGmm():gmm-init-model.cc:55) Tree has pdf-id 1 with no stats; corresponding phone list: 6 7 8 9 10 
** The warnings above about 'no stats' generally mean you have phones **
** (or groups of phones) in your phone set that had no corresponding data. **
** You should probably figure out whether something went wrong, **
** or whether your data just doesn't happen to have examples of those **
** phones. **
steps/train_deltas.sh: converting alignments from exp/mono_ali to use current tree
steps/train_deltas.sh: compiling graphs of transcripts
steps/train_deltas.sh: training pass 1
steps/train_deltas.sh: training pass 2
··················
steps/train_deltas.sh: aligning data
steps/train_deltas.sh: training pass 31
steps/train_deltas.sh: training pass 32
steps/train_deltas.sh: training pass 33
steps/train_deltas.sh: training pass 34
steps/diagnostic/analyze_alignments.sh --cmd run.pl data/lang exp/tri1
steps/diagnostic/analyze_alignments.sh: see stats in exp/tri1/log/analyze_alignments.log
382 warnings in exp/tri1/log/update.*.log
1 warnings in exp/tri1/log/questions.log
1 warnings in exp/tri1/log/build_tree.log
12 warnings in exp/tri1/log/align.*.*.log
28 warnings in exp/tri1/log/init_model.log
1 warnings in exp/tri1/log/mixup.log
exp/tri1: nj=4 align prob=-79.41 over 0.06h [retry=0.0%, fail=0.0%] states=88 gauss=1003 tree-impr=5.48
steps/train_deltas.sh: Done training system with delta+delta-delta features in exp/tri1===== TRI1 (first triphone pass) DECODING =====tree-info exp/tri1/tree 
tree-info exp/tri1/tree 
fstcomposecontext --context-size=3 --central-position=1 --read-disambig-syms=data/lang/phones/disambig.int --write-disambig-syms=data/lang/tmp/disambig_ilabels_3_1.int data/lang/tmp/ilabels_3_1 
fstisstochastic data/lang/tmp/CLG_3_1.fst 
0 -0.0423222
[info]: CLG not stochastic.
make-h-transducer --disambig-syms-out=exp/tri1/graph/disambig_tid.int --transition-scale=1.0 data/lang/tmp/ilabels_3_1 exp/tri1/tree exp/tri1/final.mdl 
fsttablecompose exp/tri1/graph/Ha.fst data/lang/tmp/CLG_3_1.fst 
fstdeterminizestar --use-log=true 
fstminimizeencoded 
fstrmsymbols exp/tri1/graph/disambig_tid.int 
fstrmepslocal 
fstisstochastic exp/tri1/graph/HCLGa.fst 
0.000384301 -0.0999308
HCLGa is not stochastic
add-self-loops --self-loop-scale=0.1 --reorder=true exp/tri1/final.mdl 
steps/decode.sh --config conf/decode.config --nj 4 --cmd run.pl exp/tri1/graph data/test exp/tri1/decode
decode.sh: feature type is delta
steps/diagnostic/analyze_lats.sh --cmd run.pl exp/tri1/graph exp/tri1/decode
steps/diagnostic/analyze_lats.sh: see stats in exp/tri1/decode/log/analyze_alignments.log
Overall, lattice depth (10,50,90-percentile)=(1,1,1) and mean=1.1
steps/diagnostic/analyze_lats.sh: see stats in exp/tri1/decode/log/analyze_lattice_depth_stats.log
exp/tri1/decode/wer_10
···················===== run.sh script is finished =====

tags: 目录语音编号标注脚本

笔记摘录：2018.09.01---Kaldi构建一个简单的英文数字串识别系统

p81 红蓝对抗-AWD 监控不死马垃圾包资源库

浅谈WEB安全性（前端向）

如何计算应缴个人所得税? 使用SQL，理解SQL

用AI解锁声音的奥秘——启英泰伦新成果发布交流会

将可执行jar包部署为windows服务

Flutter 中使用了plugin 在运行iOS 项目找不到.h 的问题

利用区块链等技术，加强对交通运输信用信息的归集共享和分析应用

印尼西爪哇梳邦县发生山体滑坡已经导致2人死亡

【SpringBoot笔记10】Spring中Bean的6种作用域

ARS548 ARS549RDI 80GHZ毫米波雷达达学习笔记（一)

叠氮PEG修饰二硒化钨 (N3-WSe2；azide

ATFX：黑海运粮遭俄暂停，小麦期货开盘跳涨

关于我们

最火推荐

小编推荐

联系我们

复制成功

笔记摘录：2018.09.01---Kaldi构建一个简单的英文数字串识别系统

p81 红蓝对抗-AWD 监控不死马垃圾包资源库

浅谈WEB安全性（前端向）

如何计算应缴个人所得税? 使用SQL，理解SQL

用AI解锁声音的奥秘——启英泰伦新成果发布交流会

将可执行jar包部署为windows服务

Flutter 中使用了plugin 在运行iOS 项目找不到.h 的问题

利用区块链等技术，加强对交通运输信用信息的归集共享和分析应用

印尼西爪哇梳邦县发生山体滑坡 已经导致2人死亡

【SpringBoot笔记10】Spring中Bean的6种作用域

ARS548 ARS549RDI 80GHZ毫米波雷达达学习笔记（一)

叠氮PEG修饰二硒化钨 (N3-WSe2；azide

ATFX：黑海运粮遭俄暂停，小麦期货开盘跳涨

关于我们

最火推荐

小编推荐

联系我们

复制成功

印尼西爪哇梳邦县发生山体滑坡已经导致2人死亡