Add two-pass speech recognition Android/iOS demo (#304)

This commit is contained in:
Fangjun Kuang
2023-09-12 15:40:16 +08:00
committed by GitHub
parent 8982984ea2
commit debab7c091
97 changed files with 3546 additions and 57 deletions

91
ios-swiftui/.gitignore vendored Normal file
View File

@@ -0,0 +1,91 @@
# See https://github.com/github/gitignore/blob/main/Swift.gitignore
# Xcode
#
# gitignore contributors: remember to update Global/Xcode.gitignore, Objective-C.gitignore & Swift.gitignore
## User settings
xcuserdata/
## compatibility with Xcode 8 and earlier (ignoring not required starting Xcode 9)
*.xcscmblueprint
*.xccheckout
## compatibility with Xcode 3 and earlier (ignoring not required starting Xcode 4)
build/
DerivedData/
*.moved-aside
*.pbxuser
!default.pbxuser
*.mode1v3
!default.mode1v3
*.mode2v3
!default.mode2v3
*.perspectivev3
!default.perspectivev3
## Obj-C/Swift specific
*.hmap
## App packaging
*.ipa
*.dSYM.zip
*.dSYM
## Playgrounds
timeline.xctimeline
playground.xcworkspace
# Swift Package Manager
#
# Add this line if you want to avoid checking in source code from Swift Package Manager dependencies.
# Packages/
# Package.pins
# Package.resolved
# *.xcodeproj
#
# Xcode automatically generates this directory with a .xcworkspacedata file and xcuserdata
# hence it is not needed unless you have added a package configuration file to your project
# .swiftpm
.build/
# CocoaPods
#
# We recommend against adding the Pods directory to your .gitignore. However
# you should judge for yourself, the pros and cons are mentioned at:
# https://guides.cocoapods.org/using/using-cocoapods.html#should-i-check-the-pods-directory-into-source-control
#
# Pods/
#
# Add this line if you want to avoid checking in source code from the Xcode workspace
# *.xcworkspace
# Carthage
#
# Add this line if you want to avoid checking in source code from Carthage dependencies.
# Carthage/Checkouts
Carthage/Build/
# Accio dependency management
Dependencies/
.accio/
# fastlane
#
# It is recommended to not store the screenshots in the git repo.
# Instead, use fastlane to re-generate the screenshots whenever they are needed.
# For more information about the recommended setup visit:
# https://docs.fastlane.tools/best-practices/source-control/#source-control
fastlane/report.xml
fastlane/Preview.html
fastlane/screenshots/**/*.png
fastlane/test_output
# Code Injection
#
# After new code Injection tools there's a generated folder /iOSInjectionProject
# https://github.com/johnno1962/injectionforxcode
iOSInjectionProject/

View File

@@ -0,0 +1,380 @@
// !$*UTF8*$!
{
archiveVersion = 1;
classes = {
};
objectVersion = 56;
objects = {
/* Begin PBXBuildFile section */
C9A2587D2AAEFFF100E555CA /* SherpaOnnx2PassApp.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A2587C2AAEFFF100E555CA /* SherpaOnnx2PassApp.swift */; };
C9A2587F2AAEFFF100E555CA /* ContentView.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A2587E2AAEFFF100E555CA /* ContentView.swift */; };
C9A258812AAEFFF200E555CA /* Assets.xcassets in Resources */ = {isa = PBXBuildFile; fileRef = C9A258802AAEFFF200E555CA /* Assets.xcassets */; };
C9A258842AAEFFF200E555CA /* Preview Assets.xcassets in Resources */ = {isa = PBXBuildFile; fileRef = C9A258832AAEFFF200E555CA /* Preview Assets.xcassets */; };
C9A2588E2AAF039D00E555CA /* Model.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A2588A2AAF039D00E555CA /* Model.swift */; };
C9A258902AAF039D00E555CA /* SherpaOnnxViewModel.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A2588C2AAF039D00E555CA /* SherpaOnnxViewModel.swift */; };
C9A258912AAF039D00E555CA /* Extension.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A2588D2AAF039D00E555CA /* Extension.swift */; };
C9A258932AAF057E00E555CA /* SherpaOnnx.swift in Sources */ = {isa = PBXBuildFile; fileRef = C9A258922AAF057E00E555CA /* SherpaOnnx.swift */; };
C9A258962AAF05D100E555CA /* sherpa-onnx.xcframework in Frameworks */ = {isa = PBXBuildFile; fileRef = C9A258952AAF05D100E555CA /* sherpa-onnx.xcframework */; };
C9A258982AAF05E400E555CA /* onnxruntime.xcframework in Frameworks */ = {isa = PBXBuildFile; fileRef = C9A258972AAF05E400E555CA /* onnxruntime.xcframework */; };
/* End PBXBuildFile section */
/* Begin PBXFileReference section */
C9A258792AAEFFF100E555CA /* SherpaOnnx2Pass.app */ = {isa = PBXFileReference; explicitFileType = wrapper.application; includeInIndex = 0; path = SherpaOnnx2Pass.app; sourceTree = BUILT_PRODUCTS_DIR; };
C9A2587C2AAEFFF100E555CA /* SherpaOnnx2PassApp.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = SherpaOnnx2PassApp.swift; sourceTree = "<group>"; };
C9A2587E2AAEFFF100E555CA /* ContentView.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = ContentView.swift; sourceTree = "<group>"; };
C9A258802AAEFFF200E555CA /* Assets.xcassets */ = {isa = PBXFileReference; lastKnownFileType = folder.assetcatalog; path = Assets.xcassets; sourceTree = "<group>"; };
C9A258832AAEFFF200E555CA /* Preview Assets.xcassets */ = {isa = PBXFileReference; lastKnownFileType = folder.assetcatalog; path = "Preview Assets.xcassets"; sourceTree = "<group>"; };
C9A2588A2AAF039D00E555CA /* Model.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = Model.swift; sourceTree = "<group>"; };
C9A2588C2AAF039D00E555CA /* SherpaOnnxViewModel.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = SherpaOnnxViewModel.swift; sourceTree = "<group>"; };
C9A2588D2AAF039D00E555CA /* Extension.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = Extension.swift; sourceTree = "<group>"; };
C9A258922AAF057E00E555CA /* SherpaOnnx.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; name = SherpaOnnx.swift; path = "../../../swift-api-examples/SherpaOnnx.swift"; sourceTree = "<group>"; };
C9A258952AAF05D100E555CA /* sherpa-onnx.xcframework */ = {isa = PBXFileReference; lastKnownFileType = wrapper.xcframework; name = "sherpa-onnx.xcframework"; path = "../../build-ios/sherpa-onnx.xcframework"; sourceTree = "<group>"; };
C9A258972AAF05E400E555CA /* onnxruntime.xcframework */ = {isa = PBXFileReference; lastKnownFileType = wrapper.xcframework; name = onnxruntime.xcframework; path = "../../build-ios/ios-onnxruntime/1.15.1/onnxruntime.xcframework"; sourceTree = "<group>"; };
/* End PBXFileReference section */
/* Begin PBXFrameworksBuildPhase section */
C9A258762AAEFFF100E555CA /* Frameworks */ = {
isa = PBXFrameworksBuildPhase;
buildActionMask = 2147483647;
files = (
C9A258982AAF05E400E555CA /* onnxruntime.xcframework in Frameworks */,
C9A258962AAF05D100E555CA /* sherpa-onnx.xcframework in Frameworks */,
);
runOnlyForDeploymentPostprocessing = 0;
};
/* End PBXFrameworksBuildPhase section */
/* Begin PBXGroup section */
C9A258702AAEFFF100E555CA = {
isa = PBXGroup;
children = (
C9A2587B2AAEFFF100E555CA /* SherpaOnnx2Pass */,
C9A2587A2AAEFFF100E555CA /* Products */,
C9A258942AAF05D100E555CA /* Frameworks */,
);
sourceTree = "<group>";
};
C9A2587A2AAEFFF100E555CA /* Products */ = {
isa = PBXGroup;
children = (
C9A258792AAEFFF100E555CA /* SherpaOnnx2Pass.app */,
);
name = Products;
sourceTree = "<group>";
};
C9A2587B2AAEFFF100E555CA /* SherpaOnnx2Pass */ = {
isa = PBXGroup;
children = (
C9A258922AAF057E00E555CA /* SherpaOnnx.swift */,
C9A2588D2AAF039D00E555CA /* Extension.swift */,
C9A2588A2AAF039D00E555CA /* Model.swift */,
C9A2588C2AAF039D00E555CA /* SherpaOnnxViewModel.swift */,
C9A2587C2AAEFFF100E555CA /* SherpaOnnx2PassApp.swift */,
C9A2587E2AAEFFF100E555CA /* ContentView.swift */,
C9A258802AAEFFF200E555CA /* Assets.xcassets */,
C9A258822AAEFFF200E555CA /* Preview Content */,
);
path = SherpaOnnx2Pass;
sourceTree = "<group>";
};
C9A258822AAEFFF200E555CA /* Preview Content */ = {
isa = PBXGroup;
children = (
C9A258832AAEFFF200E555CA /* Preview Assets.xcassets */,
);
path = "Preview Content";
sourceTree = "<group>";
};
C9A258942AAF05D100E555CA /* Frameworks */ = {
isa = PBXGroup;
children = (
C9A258972AAF05E400E555CA /* onnxruntime.xcframework */,
C9A258952AAF05D100E555CA /* sherpa-onnx.xcframework */,
);
name = Frameworks;
sourceTree = "<group>";
};
/* End PBXGroup section */
/* Begin PBXNativeTarget section */
C9A258782AAEFFF100E555CA /* SherpaOnnx2Pass */ = {
isa = PBXNativeTarget;
buildConfigurationList = C9A258872AAEFFF200E555CA /* Build configuration list for PBXNativeTarget "SherpaOnnx2Pass" */;
buildPhases = (
C9A258752AAEFFF100E555CA /* Sources */,
C9A258762AAEFFF100E555CA /* Frameworks */,
C9A258772AAEFFF100E555CA /* Resources */,
);
buildRules = (
);
dependencies = (
);
name = SherpaOnnx2Pass;
productName = SherpaOnnx2Pass;
productReference = C9A258792AAEFFF100E555CA /* SherpaOnnx2Pass.app */;
productType = "com.apple.product-type.application";
};
/* End PBXNativeTarget section */
/* Begin PBXProject section */
C9A258712AAEFFF100E555CA /* Project object */ = {
isa = PBXProject;
attributes = {
BuildIndependentTargetsInParallel = 1;
LastSwiftUpdateCheck = 1420;
LastUpgradeCheck = 1420;
TargetAttributes = {
C9A258782AAEFFF100E555CA = {
CreatedOnToolsVersion = 14.2;
};
};
};
buildConfigurationList = C9A258742AAEFFF100E555CA /* Build configuration list for PBXProject "SherpaOnnx2Pass" */;
compatibilityVersion = "Xcode 14.0";
developmentRegion = en;
hasScannedForEncodings = 0;
knownRegions = (
en,
Base,
);
mainGroup = C9A258702AAEFFF100E555CA;
productRefGroup = C9A2587A2AAEFFF100E555CA /* Products */;
projectDirPath = "";
projectRoot = "";
targets = (
C9A258782AAEFFF100E555CA /* SherpaOnnx2Pass */,
);
};
/* End PBXProject section */
/* Begin PBXResourcesBuildPhase section */
C9A258772AAEFFF100E555CA /* Resources */ = {
isa = PBXResourcesBuildPhase;
buildActionMask = 2147483647;
files = (
C9A258842AAEFFF200E555CA /* Preview Assets.xcassets in Resources */,
C9A258812AAEFFF200E555CA /* Assets.xcassets in Resources */,
);
runOnlyForDeploymentPostprocessing = 0;
};
/* End PBXResourcesBuildPhase section */
/* Begin PBXSourcesBuildPhase section */
C9A258752AAEFFF100E555CA /* Sources */ = {
isa = PBXSourcesBuildPhase;
buildActionMask = 2147483647;
files = (
C9A2588E2AAF039D00E555CA /* Model.swift in Sources */,
C9A258902AAF039D00E555CA /* SherpaOnnxViewModel.swift in Sources */,
C9A258912AAF039D00E555CA /* Extension.swift in Sources */,
C9A2587F2AAEFFF100E555CA /* ContentView.swift in Sources */,
C9A258932AAF057E00E555CA /* SherpaOnnx.swift in Sources */,
C9A2587D2AAEFFF100E555CA /* SherpaOnnx2PassApp.swift in Sources */,
);
runOnlyForDeploymentPostprocessing = 0;
};
/* End PBXSourcesBuildPhase section */
/* Begin XCBuildConfiguration section */
C9A258852AAEFFF200E555CA /* Debug */ = {
isa = XCBuildConfiguration;
buildSettings = {
ALWAYS_SEARCH_USER_PATHS = NO;
CLANG_ANALYZER_NONNULL = YES;
CLANG_ANALYZER_NUMBER_OBJECT_CONVERSION = YES_AGGRESSIVE;
CLANG_CXX_LANGUAGE_STANDARD = "gnu++20";
CLANG_ENABLE_MODULES = YES;
CLANG_ENABLE_OBJC_ARC = YES;
CLANG_ENABLE_OBJC_WEAK = YES;
CLANG_WARN_BLOCK_CAPTURE_AUTORELEASING = YES;
CLANG_WARN_BOOL_CONVERSION = YES;
CLANG_WARN_COMMA = YES;
CLANG_WARN_CONSTANT_CONVERSION = YES;
CLANG_WARN_DEPRECATED_OBJC_IMPLEMENTATIONS = YES;
CLANG_WARN_DIRECT_OBJC_ISA_USAGE = YES_ERROR;
CLANG_WARN_DOCUMENTATION_COMMENTS = YES;
CLANG_WARN_EMPTY_BODY = YES;
CLANG_WARN_ENUM_CONVERSION = YES;
CLANG_WARN_INFINITE_RECURSION = YES;
CLANG_WARN_INT_CONVERSION = YES;
CLANG_WARN_NON_LITERAL_NULL_CONVERSION = YES;
CLANG_WARN_OBJC_IMPLICIT_RETAIN_SELF = YES;
CLANG_WARN_OBJC_LITERAL_CONVERSION = YES;
CLANG_WARN_OBJC_ROOT_CLASS = YES_ERROR;
CLANG_WARN_QUOTED_INCLUDE_IN_FRAMEWORK_HEADER = YES;
CLANG_WARN_RANGE_LOOP_ANALYSIS = YES;
CLANG_WARN_STRICT_PROTOTYPES = YES;
CLANG_WARN_SUSPICIOUS_MOVE = YES;
CLANG_WARN_UNGUARDED_AVAILABILITY = YES_AGGRESSIVE;
CLANG_WARN_UNREACHABLE_CODE = YES;
CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
COPY_PHASE_STRIP = NO;
DEBUG_INFORMATION_FORMAT = dwarf;
ENABLE_STRICT_OBJC_MSGSEND = YES;
ENABLE_TESTABILITY = YES;
GCC_C_LANGUAGE_STANDARD = gnu11;
GCC_DYNAMIC_NO_PIC = NO;
GCC_NO_COMMON_BLOCKS = YES;
GCC_OPTIMIZATION_LEVEL = 0;
GCC_PREPROCESSOR_DEFINITIONS = (
"DEBUG=1",
"$(inherited)",
);
GCC_WARN_64_TO_32_BIT_CONVERSION = YES;
GCC_WARN_ABOUT_RETURN_TYPE = YES_ERROR;
GCC_WARN_UNDECLARED_SELECTOR = YES;
GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
GCC_WARN_UNUSED_FUNCTION = YES;
GCC_WARN_UNUSED_VARIABLE = YES;
IPHONEOS_DEPLOYMENT_TARGET = 16.2;
MTL_ENABLE_DEBUG_INFO = INCLUDE_SOURCE;
MTL_FAST_MATH = YES;
ONLY_ACTIVE_ARCH = YES;
SDKROOT = iphoneos;
SWIFT_ACTIVE_COMPILATION_CONDITIONS = DEBUG;
SWIFT_OPTIMIZATION_LEVEL = "-Onone";
};
name = Debug;
};
C9A258862AAEFFF200E555CA /* Release */ = {
isa = XCBuildConfiguration;
buildSettings = {
ALWAYS_SEARCH_USER_PATHS = NO;
CLANG_ANALYZER_NONNULL = YES;
CLANG_ANALYZER_NUMBER_OBJECT_CONVERSION = YES_AGGRESSIVE;
CLANG_CXX_LANGUAGE_STANDARD = "gnu++20";
CLANG_ENABLE_MODULES = YES;
CLANG_ENABLE_OBJC_ARC = YES;
CLANG_ENABLE_OBJC_WEAK = YES;
CLANG_WARN_BLOCK_CAPTURE_AUTORELEASING = YES;
CLANG_WARN_BOOL_CONVERSION = YES;
CLANG_WARN_COMMA = YES;
CLANG_WARN_CONSTANT_CONVERSION = YES;
CLANG_WARN_DEPRECATED_OBJC_IMPLEMENTATIONS = YES;
CLANG_WARN_DIRECT_OBJC_ISA_USAGE = YES_ERROR;
CLANG_WARN_DOCUMENTATION_COMMENTS = YES;
CLANG_WARN_EMPTY_BODY = YES;
CLANG_WARN_ENUM_CONVERSION = YES;
CLANG_WARN_INFINITE_RECURSION = YES;
CLANG_WARN_INT_CONVERSION = YES;
CLANG_WARN_NON_LITERAL_NULL_CONVERSION = YES;
CLANG_WARN_OBJC_IMPLICIT_RETAIN_SELF = YES;
CLANG_WARN_OBJC_LITERAL_CONVERSION = YES;
CLANG_WARN_OBJC_ROOT_CLASS = YES_ERROR;
CLANG_WARN_QUOTED_INCLUDE_IN_FRAMEWORK_HEADER = YES;
CLANG_WARN_RANGE_LOOP_ANALYSIS = YES;
CLANG_WARN_STRICT_PROTOTYPES = YES;
CLANG_WARN_SUSPICIOUS_MOVE = YES;
CLANG_WARN_UNGUARDED_AVAILABILITY = YES_AGGRESSIVE;
CLANG_WARN_UNREACHABLE_CODE = YES;
CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
COPY_PHASE_STRIP = NO;
DEBUG_INFORMATION_FORMAT = "dwarf-with-dsym";
ENABLE_NS_ASSERTIONS = NO;
ENABLE_STRICT_OBJC_MSGSEND = YES;
GCC_C_LANGUAGE_STANDARD = gnu11;
GCC_NO_COMMON_BLOCKS = YES;
GCC_WARN_64_TO_32_BIT_CONVERSION = YES;
GCC_WARN_ABOUT_RETURN_TYPE = YES_ERROR;
GCC_WARN_UNDECLARED_SELECTOR = YES;
GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
GCC_WARN_UNUSED_FUNCTION = YES;
GCC_WARN_UNUSED_VARIABLE = YES;
IPHONEOS_DEPLOYMENT_TARGET = 16.2;
MTL_ENABLE_DEBUG_INFO = NO;
MTL_FAST_MATH = YES;
SDKROOT = iphoneos;
SWIFT_COMPILATION_MODE = wholemodule;
SWIFT_OPTIMIZATION_LEVEL = "-O";
VALIDATE_PRODUCT = YES;
};
name = Release;
};
C9A258882AAEFFF200E555CA /* Debug */ = {
isa = XCBuildConfiguration;
buildSettings = {
ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
ASSETCATALOG_COMPILER_GLOBAL_ACCENT_COLOR_NAME = AccentColor;
CODE_SIGN_STYLE = Automatic;
CURRENT_PROJECT_VERSION = 1;
DEVELOPMENT_ASSET_PATHS = "\"SherpaOnnx2Pass/Preview Content\"";
ENABLE_PREVIEWS = YES;
GENERATE_INFOPLIST_FILE = YES;
HEADER_SEARCH_PATHS = "${PROJECT_DIR}/../../build-ios/sherpa-onnx.xcframework/Headers/";
INFOPLIST_KEY_UIApplicationSceneManifest_Generation = YES;
INFOPLIST_KEY_UIApplicationSupportsIndirectInputEvents = YES;
INFOPLIST_KEY_UILaunchScreen_Generation = YES;
INFOPLIST_KEY_UISupportedInterfaceOrientations_iPad = "UIInterfaceOrientationPortrait UIInterfaceOrientationPortraitUpsideDown UIInterfaceOrientationLandscapeLeft UIInterfaceOrientationLandscapeRight";
INFOPLIST_KEY_UISupportedInterfaceOrientations_iPhone = "UIInterfaceOrientationPortrait UIInterfaceOrientationLandscapeLeft UIInterfaceOrientationLandscapeRight";
LD_RUNPATH_SEARCH_PATHS = (
"$(inherited)",
"@executable_path/Frameworks",
);
MARKETING_VERSION = 1.0;
OTHER_LDFLAGS = "-lc++";
PRODUCT_BUNDLE_IDENTIFIER = "com.k2-fsa.org.SherpaOnnx2Pass";
PRODUCT_NAME = "$(TARGET_NAME)";
SWIFT_EMIT_LOC_STRINGS = YES;
SWIFT_OBJC_BRIDGING_HEADER = "${PROJECT_DIR}/../../swift-api-examples/SherpaOnnx-Bridging-Header.h";
SWIFT_VERSION = 5.0;
TARGETED_DEVICE_FAMILY = "1,2";
};
name = Debug;
};
C9A258892AAEFFF200E555CA /* Release */ = {
isa = XCBuildConfiguration;
buildSettings = {
ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
ASSETCATALOG_COMPILER_GLOBAL_ACCENT_COLOR_NAME = AccentColor;
CODE_SIGN_STYLE = Automatic;
CURRENT_PROJECT_VERSION = 1;
DEVELOPMENT_ASSET_PATHS = "\"SherpaOnnx2Pass/Preview Content\"";
ENABLE_PREVIEWS = YES;
GENERATE_INFOPLIST_FILE = YES;
HEADER_SEARCH_PATHS = "${PROJECT_DIR}/../../build-ios/sherpa-onnx.xcframework/Headers/";
INFOPLIST_KEY_UIApplicationSceneManifest_Generation = YES;
INFOPLIST_KEY_UIApplicationSupportsIndirectInputEvents = YES;
INFOPLIST_KEY_UILaunchScreen_Generation = YES;
INFOPLIST_KEY_UISupportedInterfaceOrientations_iPad = "UIInterfaceOrientationPortrait UIInterfaceOrientationPortraitUpsideDown UIInterfaceOrientationLandscapeLeft UIInterfaceOrientationLandscapeRight";
INFOPLIST_KEY_UISupportedInterfaceOrientations_iPhone = "UIInterfaceOrientationPortrait UIInterfaceOrientationLandscapeLeft UIInterfaceOrientationLandscapeRight";
LD_RUNPATH_SEARCH_PATHS = (
"$(inherited)",
"@executable_path/Frameworks",
);
MARKETING_VERSION = 1.0;
OTHER_LDFLAGS = "-lc++";
PRODUCT_BUNDLE_IDENTIFIER = "com.k2-fsa.org.SherpaOnnx2Pass";
PRODUCT_NAME = "$(TARGET_NAME)";
SWIFT_EMIT_LOC_STRINGS = YES;
SWIFT_OBJC_BRIDGING_HEADER = "${PROJECT_DIR}/../../swift-api-examples/SherpaOnnx-Bridging-Header.h";
SWIFT_VERSION = 5.0;
TARGETED_DEVICE_FAMILY = "1,2";
};
name = Release;
};
/* End XCBuildConfiguration section */
/* Begin XCConfigurationList section */
C9A258742AAEFFF100E555CA /* Build configuration list for PBXProject "SherpaOnnx2Pass" */ = {
isa = XCConfigurationList;
buildConfigurations = (
C9A258852AAEFFF200E555CA /* Debug */,
C9A258862AAEFFF200E555CA /* Release */,
);
defaultConfigurationIsVisible = 0;
defaultConfigurationName = Release;
};
C9A258872AAEFFF200E555CA /* Build configuration list for PBXNativeTarget "SherpaOnnx2Pass" */ = {
isa = XCConfigurationList;
buildConfigurations = (
C9A258882AAEFFF200E555CA /* Debug */,
C9A258892AAEFFF200E555CA /* Release */,
);
defaultConfigurationIsVisible = 0;
defaultConfigurationName = Release;
};
/* End XCConfigurationList section */
};
rootObject = C9A258712AAEFFF100E555CA /* Project object */;
}

View File

@@ -0,0 +1,7 @@
<?xml version="1.0" encoding="UTF-8"?>
<Workspace
version = "1.0">
<FileRef
location = "self:">
</FileRef>
</Workspace>

View File

@@ -0,0 +1,8 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>IDEDidComputeMac32BitWarning</key>
<true/>
</dict>
</plist>

View File

@@ -0,0 +1,11 @@
{
"colors" : [
{
"idiom" : "universal"
}
],
"info" : {
"author" : "xcode",
"version" : 1
}
}

View File

@@ -0,0 +1,14 @@
{
"images" : [
{
"filename" : "k2-1024x1024.png",
"idiom" : "universal",
"platform" : "ios",
"size" : "1024x1024"
}
],
"info" : {
"author" : "xcode",
"version" : 1
}
}

Binary file not shown.

After

Width:  |  Height:  |  Size: 411 KiB

View File

@@ -0,0 +1,6 @@
{
"info" : {
"author" : "xcode",
"version" : 1
}
}

View File

@@ -0,0 +1,46 @@
//
// ContentView.swift
// SherpaOnnx2Pass
//
// Created by fangjun on 2023/9/11.
//
import SwiftUI
struct ContentView: View {
@StateObject var sherpaOnnxVM = SherpaOnnxViewModel()
var body: some View {
VStack {
Text("ASR with Next-gen Kaldi")
.font(.title)
if sherpaOnnxVM.status == .stop {
Text("See https://github.com/k2-fsa/sherpa-onnx")
Text("Press the Start button to run!")
}
ScrollView(.vertical, showsIndicators: true) {
HStack {
Text(sherpaOnnxVM.subtitles)
Spacer()
}
}
Spacer()
Button {
toggleRecorder()
} label: {
Text(sherpaOnnxVM.status == .stop ? "Start" : "Stop")
}
}
.padding()
}
private func toggleRecorder() {
sherpaOnnxVM.toggleRecorder()
}
}
struct ContentView_Previews: PreviewProvider {
static var previews: some View {
ContentView()
}
}

View File

@@ -0,0 +1,20 @@
//
// Extension.swift
// SherpaOnnx
//
// Created by knight on 2023/4/5.
//
import AVFoundation
extension AudioBuffer {
func array() -> [Float] {
return Array(UnsafeBufferPointer(self))
}
}
extension AVAudioPCMBuffer {
func array() -> [Float] {
return self.audioBufferList.pointee.mBuffers.array()
}
}

View File

@@ -0,0 +1,134 @@
import Foundation
func getResource(_ forResource: String, _ ofType: String) -> String {
let path = Bundle.main.path(forResource: forResource, ofType: ofType)
precondition(
path != nil,
"\(forResource).\(ofType) does not exist!\n" + "Remember to change \n"
+ " Build Phases -> Copy Bundle Resources\n" + "to add it!"
)
return path!
}
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to download pre-trained models
/// sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20 (Bilingual, Chinese + English)
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/zipformer-transducer-models.html
func getBilingualStreamingZhEnZipformer20230220() -> SherpaOnnxOnlineModelConfig {
let encoder = getResource("encoder-epoch-99-avg-1.int8", "onnx")
let decoder = getResource("decoder-epoch-99-avg-1", "onnx")
let joiner = getResource("joiner-epoch-99-avg-1.int8", "onnx")
let tokens = getResource("tokens", "txt")
return sherpaOnnxOnlineModelConfig(
tokens: tokens,
transducer: sherpaOnnxOnlineTransducerModelConfig(
encoder: encoder,
decoder: decoder,
joiner: joiner),
numThreads: 1,
modelType: "zipformer"
)
}
/// csukuangfj/sherpa-onnx-streaming-zipformer-zh-14M-2023-02-23 (Chinese)
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/online-transducer/zipformer-transducer-models.html#csukuangfj-sherpa-onnx-streaming-zipformer-zh-14m-2023-02-23-chinese
func getStreamingZh14MZipformer20230223() -> SherpaOnnxOnlineModelConfig {
let encoder = getResource("encoder-epoch-99-avg-1.int8", "onnx")
let decoder = getResource("decoder-epoch-99-avg-1", "onnx")
let joiner = getResource("joiner-epoch-99-avg-1.int8", "onnx")
let tokens = getResource("tokens", "txt")
return sherpaOnnxOnlineModelConfig(
tokens: tokens,
transducer: sherpaOnnxOnlineTransducerModelConfig(
encoder: encoder,
decoder: decoder,
joiner: joiner),
numThreads: 1,
modelType: "zipformer"
)
}
/// csukuangfj/sherpa-onnx-streaming-zipformer-en-20M-2023-02-17 (English)
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/online-transducer/zipformer-transducer-models.html#csukuangfj-sherpa-onnx-streaming-zipformer-en-20m-2023-02-17-english
func getStreamingEn20MZipformer20230217() -> SherpaOnnxOnlineModelConfig {
let encoder = getResource("encoder-epoch-99-avg-1.int8", "onnx")
let decoder = getResource("decoder-epoch-99-avg-1", "onnx")
let joiner = getResource("joiner-epoch-99-avg-1", "onnx")
let tokens = getResource("tokens", "txt")
return sherpaOnnxOnlineModelConfig(
tokens: tokens,
transducer: sherpaOnnxOnlineTransducerModelConfig(
encoder: encoder,
decoder: decoder,
joiner: joiner),
numThreads: 1,
modelType: "zipformer"
)
}
/// ========================================
/// Non-streaming models
/// ========================================
/// csukuangfj/sherpa-onnx-paraformer-zh-2023-03-28 (Chinese)
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-paraformer/paraformer-models.html#csukuangfj-sherpa-onnx-paraformer-zh-2023-03-28-chinese
func getNonStreamingZhParaformer20230328() -> SherpaOnnxOfflineModelConfig {
let model = getResource("model.int8", "onnx")
let tokens = getResource("paraformer-tokens", "txt")
return sherpaOnnxOfflineModelConfig(
tokens: tokens,
paraformer: sherpaOnnxOfflineParaformerModelConfig(
model: model),
numThreads: 1,
modelType: "paraformer"
)
}
// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/whisper/tiny.en.html#tiny-en
// English, int8 encoder and decoder
func getNonStreamingWhisperTinyEn() -> SherpaOnnxOfflineModelConfig {
let encoder = getResource("tiny.en-encoder.int8", "onnx")
let decoder = getResource("tiny.en-decoder.int8", "onnx")
let tokens = getResource("tiny.en-tokens", "txt")
return sherpaOnnxOfflineModelConfig(
tokens: tokens,
whisper: sherpaOnnxOfflineWhisperModelConfig(
encoder: encoder,
decoder: decoder
),
numThreads: 1,
modelType: "whisper"
)
}
// icefall-asr-multidataset-pruned_transducer_stateless7-2023-05-04 (English)
// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/offline-transducer/zipformer-transducer-models.html#icefall-asr-multidataset-pruned-transducer-stateless7-2023-05-04-english
func getNonStreamingEnZipformer20230504() -> SherpaOnnxOfflineModelConfig {
let encoder = getResource("encoder-epoch-30-avg-4.int8", "onnx")
let decoder = getResource("decoder-epoch-30-avg-4", "onnx")
let joiner = getResource("joiner-epoch-30-avg-4", "onnx")
let tokens = getResource("non-streaming-zipformer-tokens", "txt")
return sherpaOnnxOfflineModelConfig(
tokens: tokens,
transducer: sherpaOnnxOfflineTransducerModelConfig(
encoder: encoder,
decoder: decoder,
joiner: joiner),
numThreads: 1,
modelType: "zipformer"
)
}
/// Please refer to
/// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
/// to add more models if you need

View File

@@ -0,0 +1,6 @@
{
"info" : {
"author" : "xcode",
"version" : 1
}
}

View File

@@ -0,0 +1,17 @@
//
// SherpaOnnx2PassApp.swift
// SherpaOnnx2Pass
//
// Created by fangjun on 2023/9/11.
//
import SwiftUI
@main
struct SherpaOnnx2PassApp: App {
var body: some Scene {
WindowGroup {
ContentView()
}
}
}

View File

@@ -0,0 +1,252 @@
//
// SherpaOnnxViewModel.swift
// SherpaOnnx
//
// Created by knight on 2023/4/5.
//
import Foundation
import AVFoundation
enum Status {
case stop
case recording
}
class SherpaOnnxViewModel: ObservableObject {
@Published var status: Status = .stop
@Published var subtitles: String = ""
var sentences: [String] = []
var samplesBuffer = [[Float]] ()
var audioEngine: AVAudioEngine? = nil
var recognizer: SherpaOnnxRecognizer! = nil
var offlineRecognizer: SherpaOnnxOfflineRecognizer! = nil
var lastSentence: String = ""
// let maxSentence: Int = 10 // for Chinese
let maxSentence: Int = 6 // for English
var results: String {
if sentences.isEmpty && lastSentence.isEmpty {
return ""
}
if sentences.isEmpty {
return "0: \(lastSentence.lowercased())"
}
let start = max(sentences.count - maxSentence, 0)
if lastSentence.isEmpty {
return sentences.enumerated().map { (index, s) in "\(index): \(s.lowercased())" }[start...]
.joined(separator: "\n")
} else {
return sentences.enumerated().map { (index, s) in "\(index): \(s.lowercased())" }[start...]
.joined(separator: "\n") + "\n\(sentences.count): \(lastSentence.lowercased())"
}
}
func updateLabel() {
DispatchQueue.main.async {
self.subtitles = self.results
}
}
init() {
initRecognizer()
initOfflineRecognizer()
initRecorder()
}
private func initRecognizer() {
// Please select one model that is best suitable for you.
//
// You can also modify Model.swift to add new pre-trained models from
// https://k2-fsa.github.io/sherpa/onnx/pretrained_models/index.html
// let modelConfig = getBilingualStreamingZhEnZipformer20230220()
/* let modelConfig = getStreamingZh14MZipformer20230223() */
let modelConfig = getStreamingEn20MZipformer20230217()
let featConfig = sherpaOnnxFeatureConfig(
sampleRate: 16000,
featureDim: 80)
var config = sherpaOnnxOnlineRecognizerConfig(
featConfig: featConfig,
modelConfig: modelConfig,
enableEndpoint: true,
rule1MinTrailingSilence: 2.4,
// rule2MinTrailingSilence: 1.2, // for Chinese
rule2MinTrailingSilence: 0.5, // for English
rule3MinUtteranceLength: 30,
decodingMethod: "greedy_search",
maxActivePaths: 4
)
recognizer = SherpaOnnxRecognizer(config: &config)
}
private func initOfflineRecognizer() {
// let modelConfig = getNonStreamingZhParaformer20230328()
let modelConfig = getNonStreamingWhisperTinyEn()
// let modelConfig = getNonStreamingEnZipformer20230504()
let featConfig = sherpaOnnxFeatureConfig(
sampleRate: 16000,
featureDim: 80)
var config = sherpaOnnxOfflineRecognizerConfig(
featConfig: featConfig,
modelConfig: modelConfig,
decodingMethod: "greedy_search",
maxActivePaths: 4
)
offlineRecognizer = SherpaOnnxOfflineRecognizer(config: &config)
}
private func initRecorder() {
print("init recorder")
audioEngine = AVAudioEngine()
let inputNode = self.audioEngine?.inputNode
let bus = 0
let inputFormat = inputNode?.outputFormat(forBus: bus)
let outputFormat = AVAudioFormat(
commonFormat: .pcmFormatFloat32,
sampleRate: 16000, channels: 1,
interleaved: false)!
let converter = AVAudioConverter(from: inputFormat!, to: outputFormat)!
inputNode!.installTap(
onBus: bus,
bufferSize: 1024,
format: inputFormat
) {
(buffer: AVAudioPCMBuffer, when: AVAudioTime) in
var newBufferAvailable = true
let inputCallback: AVAudioConverterInputBlock = {
inNumPackets, outStatus in
if newBufferAvailable {
outStatus.pointee = .haveData
newBufferAvailable = false
return buffer
} else {
outStatus.pointee = .noDataNow
return nil
}
}
let convertedBuffer = AVAudioPCMBuffer(
pcmFormat: outputFormat,
frameCapacity:
AVAudioFrameCount(outputFormat.sampleRate)
* buffer.frameLength
/ AVAudioFrameCount(buffer.format.sampleRate))!
var error: NSError?
let _ = converter.convert(
to: convertedBuffer,
error: &error, withInputFrom: inputCallback)
// TODO(fangjun): Handle status != haveData
let array = convertedBuffer.array()
if !array.isEmpty {
self.samplesBuffer.append(array)
self.recognizer.acceptWaveform(samples: array)
while (self.recognizer.isReady()){
self.recognizer.decode()
}
let isEndpoint = self.recognizer.isEndpoint()
let text = self.recognizer.getResult().text
if !text.isEmpty && self.lastSentence != text {
self.lastSentence = text
self.updateLabel()
print(text)
}
if isEndpoint{
if !text.isEmpty {
// Invoke offline recognizer
var numSamples: Int = 0
for a in self.samplesBuffer {
numSamples += a.count
}
var samples: [Float] = Array(repeating: 0, count: numSamples)
var i = 0
for a in self.samplesBuffer {
for s in a {
samples[i] = s
i += 1
}
}
// let num = 12000 // For Chinese
let num = 10000 // For English
self.lastSentence = self.offlineRecognizer.decode(samples: Array(samples[0..<samples.count-num])).text
let tmp = self.lastSentence
self.lastSentence = ""
self.sentences.append(tmp)
self.updateLabel()
i = 0
if samples.count > num {
i = samples.count - num
}
var tail: [Float] = Array(repeating: 0, count: samples.count - i)
for k in 0 ... samples.count - i - 1 {
tail[k] = samples[i+k];
}
self.samplesBuffer = [[Float]]()
self.samplesBuffer.append(tail)
} else {
self.samplesBuffer = [[Float]]()
}
self.recognizer.reset()
}
}
}
}
public func toggleRecorder() {
if status == .stop {
startRecorder()
status = .recording
} else {
stopRecorder()
status = .stop
}
}
private func startRecorder() {
lastSentence = ""
sentences = []
samplesBuffer = [[Float]] ()
updateLabel()
do {
try self.audioEngine?.start()
} catch let error as NSError {
print("Got an error starting audioEngine: \(error.domain), \(error)")
}
print("started")
}
private func stopRecorder() {
audioEngine?.stop()
print("stopped")
}
}

View File

@@ -0,0 +1 @@
../../SherpaOnnx/SherpaOnnx/k2-1024x1024.png