EngineX-Iluvatar/enginex_bi_series-sherpa-onnx
Archived
16
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

Add inverse text normalization for online ASR (#1020)

Browse Source
This commit is contained in:
Fangjun Kuang
2024-06-17 18:39:23 +08:00
committed by GitHub
parent 6e09933d99
commit 349d957da2
12 changed files with 390 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
sherpa-onnx/csrc/online-recognizer-ctc-impl.h
View File

@@ -68,7 +68,8 @@ static OnlineRecognizerResult Convert(const OnlineCtcDecoderResult &src,
class OnlineRecognizerCtcImpl : public OnlineRecognizerImpl {
public:
explicit OnlineRecognizerCtcImpl(const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(config),
config_(config),
model_(OnlineCtcModel::Create(config.model_config)),
sym_(config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -84,7 +85,8 @@ class OnlineRecognizerCtcImpl : public OnlineRecognizerImpl {
#if __ANDROID_API__ >= 9
explicit OnlineRecognizerCtcImpl(AAssetManager *mgr,
const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(mgr, config),
config_(config),
model_(OnlineCtcModel::Create(mgr, config.model_config)),
sym_(mgr, config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -182,8 +184,10 @@ class OnlineRecognizerCtcImpl : public OnlineRecognizerImpl {
// TODO(fangjun): Remember to change these constants if needed
int32_t frame_shift_ms = 10;
int32_t subsampling_factor = 4;
return Convert(decoder_result, sym_, frame_shift_ms, subsampling_factor,
s->GetCurrentSegment(), s->GetNumFramesSinceStart());
auto r = Convert(decoder_result, sym_, frame_shift_ms, subsampling_factor,
s->GetCurrentSegment(), s->GetNumFramesSinceStart());
r.text = ApplyInverseTextNormalization(r.text);
return r;
}
bool IsEndpoint(OnlineStream *s) const override {

117
sherpa-onnx/csrc/online-recognizer-impl.cc
View File

@@ -4,11 +4,22 @@
#include "sherpa-onnx/csrc/online-recognizer-impl.h"
#if __ANDROID_API__ >= 9
#include <strstream>
#include "android/asset_manager.h"
#include "android/asset_manager_jni.h"
#endif
#include "fst/extensions/far/far.h"
#include "kaldifst/csrc/kaldi-fst-io.h"
#include "sherpa-onnx/csrc/macros.h"
#include "sherpa-onnx/csrc/online-recognizer-ctc-impl.h"
#include "sherpa-onnx/csrc/online-recognizer-paraformer-impl.h"
#include "sherpa-onnx/csrc/online-recognizer-transducer-impl.h"
#include "sherpa-onnx/csrc/online-recognizer-transducer-nemo-impl.h"
#include "sherpa-onnx/csrc/onnx-utils.h"
#include "sherpa-onnx/csrc/text-utils.h"
namespace sherpa_onnx {
@@ -78,4 +89,110 @@ std::unique_ptr<OnlineRecognizerImpl> OnlineRecognizerImpl::Create(
}
#endif
OnlineRecognizerImpl::OnlineRecognizerImpl(const OnlineRecognizerConfig &config)
: config_(config) {
if (!config.rule_fsts.empty()) {
std::vector<std::string> files;
SplitStringToVector(config.rule_fsts, ",", false, &files);
itn_list_.reserve(files.size());
for (const auto &f : files) {
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("rule fst: %s", f.c_str());
}
itn_list_.push_back(std::make_unique<kaldifst::TextNormalizer>(f));
}
}
if (!config.rule_fars.empty()) {
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("Loading FST archives");
}
std::vector<std::string> files;
SplitStringToVector(config.rule_fars, ",", false, &files);
itn_list_.reserve(files.size() + itn_list_.size());
for (const auto &f : files) {
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("rule far: %s", f.c_str());
}
std::unique_ptr<fst::FarReader<fst::StdArc>> reader(
fst::FarReader<fst::StdArc>::Open(f));
for (; !reader->Done(); reader->Next()) {
std::unique_ptr<fst::StdConstFst> r(
fst::CastOrConvertToConstFst(reader->GetFst()->Copy()));
itn_list_.push_back(
std::make_unique<kaldifst::TextNormalizer>(std::move(r)));
}
}
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("FST archives loaded!");
}
}
}
#if __ANDROID_API__ >= 9
OnlineRecognizerImpl::OnlineRecognizerImpl(AAssetManager *mgr,
const OnlineRecognizerConfig &config)
: config_(config) {
if (!config.rule_fsts.empty()) {
std::vector<std::string> files;
SplitStringToVector(config.rule_fsts, ",", false, &files);
itn_list_.reserve(files.size());
for (const auto &f : files) {
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("rule fst: %s", f.c_str());
}
auto buf = ReadFile(mgr, f);
std::istrstream is(buf.data(), buf.size());
itn_list_.push_back(std::make_unique<kaldifst::TextNormalizer>(is));
}
}
if (!config.rule_fars.empty()) {
std::vector<std::string> files;
SplitStringToVector(config.rule_fars, ",", false, &files);
itn_list_.reserve(files.size() + itn_list_.size());
for (const auto &f : files) {
if (config.model_config.debug) {
SHERPA_ONNX_LOGE("rule far: %s", f.c_str());
}
auto buf = ReadFile(mgr, f);
std::unique_ptr<std::istream> s(
new std::istrstream(buf.data(), buf.size()));
std::unique_ptr<fst::FarReader<fst::StdArc>> reader(
fst::FarReader<fst::StdArc>::Open(std::move(s)));
for (; !reader->Done(); reader->Next()) {
std::unique_ptr<fst::StdConstFst> r(
fst::CastOrConvertToConstFst(reader->GetFst()->Copy()));
itn_list_.push_back(
std::make_unique<kaldifst::TextNormalizer>(std::move(r)));
} // for (; !reader->Done(); reader->Next())
} // for (const auto &f : files)
} // if (!config.rule_fars.empty())
}
#endif
std::string OnlineRecognizerImpl::ApplyInverseTextNormalization(
std::string text) const {
if (!itn_list_.empty()) {
for (const auto &tn : itn_list_) {
text = tn->Normalize(text);
if (config_.model_config.debug) {
SHERPA_ONNX_LOGE("After inverse text normalization: %s", text.c_str());
}
}
}
return text;
}
} // namespace sherpa_onnx

20
sherpa-onnx/csrc/online-recognizer-impl.h
View File

@@ -9,6 +9,12 @@
#include <string>
#include <vector>
#if __ANDROID_API__ >= 9
#include "android/asset_manager.h"
#include "android/asset_manager_jni.h"
#endif
#include "kaldifst/csrc/text-normalizer.h"
#include "sherpa-onnx/csrc/macros.h"
#include "sherpa-onnx/csrc/online-recognizer.h"
#include "sherpa-onnx/csrc/online-stream.h"
@@ -17,10 +23,15 @@ namespace sherpa_onnx {
class OnlineRecognizerImpl {
public:
explicit OnlineRecognizerImpl(const OnlineRecognizerConfig &config);
static std::unique_ptr<OnlineRecognizerImpl> Create(
const OnlineRecognizerConfig &config);
#if __ANDROID_API__ >= 9
OnlineRecognizerImpl(AAssetManager *mgr,
const OnlineRecognizerConfig &config);
static std::unique_ptr<OnlineRecognizerImpl> Create(
AAssetManager *mgr, const OnlineRecognizerConfig &config);
#endif
@@ -50,6 +61,15 @@ class OnlineRecognizerImpl {
virtual bool IsEndpoint(OnlineStream *s) const = 0;
virtual void Reset(OnlineStream *s) const = 0;
std::string ApplyInverseTextNormalization(std::string text) const;
private:
OnlineRecognizerConfig config_;
// for inverse text normalization. Used only if
// config.rule_fsts is not empty or
// config.rule_fars is not empty
std::vector<std::unique_ptr<kaldifst::TextNormalizer>> itn_list_;
};
} // namespace sherpa_onnx

10
sherpa-onnx/csrc/online-recognizer-paraformer-impl.h
View File

@@ -96,7 +96,8 @@ static void Scale(const float *x, int32_t n, float scale, float *y) {
class OnlineRecognizerParaformerImpl : public OnlineRecognizerImpl {
public:
explicit OnlineRecognizerParaformerImpl(const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(config),
config_(config),
model_(config.model_config),
sym_(config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -116,7 +117,8 @@ class OnlineRecognizerParaformerImpl : public OnlineRecognizerImpl {
#if __ANDROID_API__ >= 9
explicit OnlineRecognizerParaformerImpl(AAssetManager *mgr,
const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(mgr, config),
config_(config),
model_(mgr, config.model_config),
sym_(mgr, config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -160,7 +162,9 @@ class OnlineRecognizerParaformerImpl : public OnlineRecognizerImpl {
OnlineRecognizerResult GetResult(OnlineStream *s) const override {
auto decoder_result = s->GetParaformerResult();
return Convert(decoder_result, sym_);
auto r = Convert(decoder_result, sym_);
r.text = ApplyInverseTextNormalization(r.text);
return r;
}
bool IsEndpoint(OnlineStream *s) const override {

12
sherpa-onnx/csrc/online-recognizer-transducer-impl.h
View File

@@ -80,7 +80,8 @@ OnlineRecognizerResult Convert(const OnlineTransducerDecoderResult &src,
class OnlineRecognizerTransducerImpl : public OnlineRecognizerImpl {
public:
explicit OnlineRecognizerTransducerImpl(const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(config),
config_(config),
model_(OnlineTransducerModel::Create(config.model_config)),
sym_(config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -124,7 +125,8 @@ class OnlineRecognizerTransducerImpl : public OnlineRecognizerImpl {
#if __ANDROID_API__ >= 9
explicit OnlineRecognizerTransducerImpl(AAssetManager *mgr,
const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(mgr, config),
config_(config),
model_(OnlineTransducerModel::Create(mgr, config.model_config)),
sym_(mgr, config.model_config.tokens),
endpoint_(config_.endpoint_config) {
@@ -332,8 +334,10 @@ class OnlineRecognizerTransducerImpl : public OnlineRecognizerImpl {
// TODO(fangjun): Remember to change these constants if needed
int32_t frame_shift_ms = 10;
int32_t subsampling_factor = 4;
return Convert(decoder_result, sym_, frame_shift_ms, subsampling_factor,
s->GetCurrentSegment(), s->GetNumFramesSinceStart());
auto r = Convert(decoder_result, sym_, frame_shift_ms, subsampling_factor,
s->GetCurrentSegment(), s->GetNumFramesSinceStart());
r.text = ApplyInverseTextNormalization(std::move(r.text));
return r;
}
bool IsEndpoint(OnlineStream *s) const override {

14
sherpa-onnx/csrc/online-recognizer-transducer-nemo-impl.h
View File

@@ -42,7 +42,8 @@ class OnlineRecognizerTransducerNeMoImpl : public OnlineRecognizerImpl {
public:
explicit OnlineRecognizerTransducerNeMoImpl(
const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(config),
config_(config),
symbol_table_(config.model_config.tokens),
endpoint_(config_.endpoint_config),
model_(
@@ -61,7 +62,8 @@ class OnlineRecognizerTransducerNeMoImpl : public OnlineRecognizerImpl {
#if __ANDROID_API__ >= 9
explicit OnlineRecognizerTransducerNeMoImpl(
AAssetManager *mgr, const OnlineRecognizerConfig &config)
: config_(config),
: OnlineRecognizerImpl(mgr, config),
config_(config),
symbol_table_(mgr, config.model_config.tokens),
endpoint_(config_.endpoint_config),
model_(std::make_unique<OnlineTransducerNeMoModel>(
@@ -94,9 +96,11 @@ class OnlineRecognizerTransducerNeMoImpl : public OnlineRecognizerImpl {
// TODO(fangjun): Remember to change these constants if needed
int32_t frame_shift_ms = 10;
int32_t subsampling_factor = model_->SubsamplingFactor();
return Convert(s->GetResult(), symbol_table_, frame_shift_ms,
subsampling_factor, s->GetCurrentSegment(),
s->GetNumFramesSinceStart());
auto r = Convert(s->GetResult(), symbol_table_, frame_shift_ms,
subsampling_factor, s->GetCurrentSegment(),
s->GetNumFramesSinceStart());
r.text = ApplyInverseTextNormalization(std::move(r.text));
return r;
}
bool IsEndpoint(OnlineStream *s) const override {

43
sherpa-onnx/csrc/online-recognizer.cc
View File

@@ -14,7 +14,9 @@
#include <utility>
#include <vector>
#include "sherpa-onnx/csrc/file-utils.h"
#include "sherpa-onnx/csrc/online-recognizer-impl.h"
#include "sherpa-onnx/csrc/text-utils.h"
namespace sherpa_onnx {
@@ -100,6 +102,15 @@ void OnlineRecognizerConfig::Register(ParseOptions *po) {
"now support greedy_search and modified_beam_search.");
po->Register("temperature-scale", &temperature_scale,
"Temperature scale for confidence computation in decoding.");
po->Register(
"rule-fsts", &rule_fsts,
"If not empty, it specifies fsts for inverse text normalization. "
"If there are multiple fsts, they are separated by a comma.");
po->Register(
"rule-fars", &rule_fars,
"If not empty, it specifies fst archives for inverse text normalization. "
"If there are multiple archives, they are separated by a comma.");
}
bool OnlineRecognizerConfig::Validate() const {
@@ -129,6 +140,34 @@ bool OnlineRecognizerConfig::Validate() const {
return false;
}
if (!hotwords_file.empty() && !FileExists(hotwords_file)) {
SHERPA_ONNX_LOGE("--hotwords-file: '%s' does not exist",
hotwords_file.c_str());
return false;
}
if (!rule_fsts.empty()) {
std::vector<std::string> files;
SplitStringToVector(rule_fsts, ",", false, &files);
for (const auto &f : files) {
if (!FileExists(f)) {
SHERPA_ONNX_LOGE("Rule fst '%s' does not exist. ", f.c_str());
return false;
}
}
}
if (!rule_fars.empty()) {
std::vector<std::string> files;
SplitStringToVector(rule_fars, ",", false, &files);
for (const auto &f : files) {
if (!FileExists(f)) {
SHERPA_ONNX_LOGE("Rule far '%s' does not exist. ", f.c_str());
return false;
}
}
}
return model_config.Validate();
}
@@ -147,7 +186,9 @@ std::string OnlineRecognizerConfig::ToString() const {
os << "hotwords_file=\"" << hotwords_file << "\", ";
os << "decoding_method=\"" << decoding_method << "\", ";
os << "blank_penalty=" << blank_penalty << ", ";
os << "temperature_scale=" << temperature_scale << ")";
os << "temperature_scale=" << temperature_scale << ", ";
os << "rule_fsts=\"" << rule_fsts << "\", ";
os << "rule_fars=\"" << rule_fars << "\")";
return os.str();
}

13
sherpa-onnx/csrc/online-recognizer.h
View File

@@ -100,6 +100,12 @@ struct OnlineRecognizerConfig {
float temperature_scale = 2.0;
// If there are multiple rules, they are applied from left to right.
std::string rule_fsts;
// If there are multiple FST archives, they are applied from left to right.
std::string rule_fars;
OnlineRecognizerConfig() = default;
OnlineRecognizerConfig(
@@ -109,7 +115,8 @@ struct OnlineRecognizerConfig {
const OnlineCtcFstDecoderConfig &ctc_fst_decoder_config,
bool enable_endpoint, const std::string &decoding_method,
int32_t max_active_paths, const std::string &hotwords_file,
float hotwords_score, float blank_penalty, float temperature_scale)
float hotwords_score, float blank_penalty, float temperature_scale,
const std::string &rule_fsts, const std::string &rule_fars)
: feat_config(feat_config),
model_config(model_config),
lm_config(lm_config),
@@ -121,7 +128,9 @@ struct OnlineRecognizerConfig {
hotwords_file(hotwords_file),
hotwords_score(hotwords_score),
blank_penalty(blank_penalty),
temperature_scale(temperature_scale) {}
temperature_scale(temperature_scale),
rule_fsts(rule_fsts),
rule_fars(rule_fars) {}
void Register(ParseOptions *po);
bool Validate() const;