文章目录 [隐藏]
本文记录于2020年。各种库、软件、系统都在更新，参考时难免有差池。欢迎讨论。
需求：将kaldi编译为 Android8.0-api26、ARMv8 使用的so动态库文件。
环境1：XUbuntu20.04.3，python3.8.10、android_ndk_r20b、git、apt。（建议的编译环境）
环境2：Mac10.14.3，python3、android-ndk、git、brew。
本文参考：
https://www.jianshu.com/p/a896bc4c3c14
https://www.jianshu.com/p/905214cedf97
http://jcsilva.github.io/2017/03/18/compile-kaldi-android/
https://medium.com/swlh/compile-kaldi-for-64-bit-android-on-ubuntu-18-70967eb3a308
安装一些必须的程序：
● linux指令：
|
1 2 3 |
sudo apt-get install cmake g++ automake autoconf git sox gfortran libtool subversion python2.7 zlib1g-dev python3-distutils |
● macos指令：
|
1 2 3 |
brew install XXX |
gfortran可以从 https://github.com/fxcoudert/gfortran-for-macOS/releases 下载
也可以根据下文步骤说明安装。
1.准备toolchain
下载地址：https://developer.android.google.cn/ndk/downloads?hl=zh_cn ，
新版(r23b)的dmg解压出来是个app，右键显示包内容，将“NDK”目录提取出来即可。
本例为r20b，解压到用户目录下 ~/apps/android-ndk-r20b。
不要使用ndk内置的 ~/apps/android-ndk-r20b/toolchains，通过 make_standalone_toolchain.py 手动创建：
|
1 2 3 4 |
python3 ~/apps/android-ndk-r20b/build/tools/make_standalone_toolchain.py \ --arch arm64 \ # 芯片位 --api 26 \ # AndroidAPI版本 --install-dir ~/apps/android-ndk-r20b/arm64-toolchain # 生成目录 |
终端输入以下3句，或配置到环境变量：
|
1 2 3 |
export ANDROID_NDK_HOME=~/apps/android-ndk-r20b export ANDROID_TOOLCHAIN_HOME=~/apps/android-ndk-r20b/arm64-toolchain export PATH=$ANDROID_NDK_HOME:$ANDROID_TOOLCHAIN_HOME/bin:$PATH:. |
linux可以配置到 ~/.bashrc，macos可以配置到 ~/.bash_profile。
2.编译openblas
下载地址：https://github.com/xianyi/OpenBLAS/releases
本例为 OpenBLAS-0.3.18.tar.gz，解压到用户目录下 ~/apps/openblas_0.3.18。
编译参考 https://github.com/xianyi/OpenBLAS/wiki/How-to-build-OpenBLAS-for-Android
|
1 2 3 4 5 6 7 8 9 10 11 |
cd ~/apps/openblas_0.3.18/ make clean make \ TARGET=ARMV8 \ BINARY=64 \ ONLY_CBLAS=1 \ CC="$ANDROID_TOOLCHAIN_HOME"/bin/aarch64-linux-android26-clang \ AR="$ANDROID_TOOLCHAIN_HOME"/bin/aarch64-linux-android-ar \ # r23b版ndk没有 ***_ar HOSTCC=gcc \ ARM_SOFTFP_ABI=1 \ -j4 |
|
1 2 3 4 5 6 7 8 9 10 11 |
echo OK. OK. rm -f linktest OpenBLAS build complete. (CBLAS) OS ... Android Architecture ... arm64 BINARY ... 64bit C compiler ... CLANG (cmd & version : Android (5220042 based on r346389c) clang version 8.0.7 (https://android.googlesource.com/toolchain/clang b55f2d4ebfd35bf643d27dbca1bb228957008617) (https://android.googlesource.com/toolchain/llvm 3c393fe7a7e13b0fba4ac75a01aa683d7a5b11cd) (based on LLVM 8.0.7svn)) -n Library Name ... libopenblas_armv8p-r0.3.18.a (Multi-threading; Max num-threads is 8) To install the library, you can run "make PREFIX=/path/to/your/installation install". |
如下指令，将编译完成的OpenBLAS安装在当前路径的install目录。
|
1 2 3 |
make install NO_SHARED=1 PREFIX=`pwd`/install |
|
1 2 3 4 5 6 7 8 9 10 |
/Library/Developer/CommandLineTools/usr/bin/make -j 8 -f Makefile.install install Generating openblas_config.h in ~/apps/openblas_0.3.18/install/include Generating f77blas.h in ~/apps/openblas_0.3.18/install/include Generating cblas.h in ~/apps/openblas_0.3.18/install/include Copying LAPACKE header files to ~/apps/openblas_0.3.18/install/include Copying the static library to ~/apps/openblas_0.3.18/install/lib Generating openblas.pc in ~/apps/openblas_0.3.18/install/lib/pkgconfig Generating OpenBLASConfig.cmake in ~/apps/openblas_0.3.18/install/lib/cmake/openblas Generating OpenBLASConfigVersion.cmake in ~/apps/openblas_0.3.18/install/lib/cmake/openblas Install OK! |
3.编译clapack
LAPACK的全称是Linear Algebra PACKage，线性代数库，用Fortran语言开发。下载地址：https://www.netlib.org/lapack/ 。CLAPACK是LAPACK的C语言接口。
下载地址：https://github.com/simonlynen/android_libs/tree/master/lapack/jni 直接使用，
也可以从 https://www.netlib.org/clapack 下载“clapack.tgz”,参照上面的添加mk文件进行修改，本例为github版本3.2.1。
备选链接：https://pan.baidu.com/s/1RU-Md0Z5iK_2WW5-ITJkXA （提取码: jn4i）
或 https://download.csdn.net/download/Vigiles/12245335
解压到用户目录下 ~/apps/clapack_3.2.1。
修改 Application.mk 文件：
|
1 2 3 4 5 6 |
APP_STL := c++_static # c++_static或c++_shared APP_CPPFLAGS := -frtti -fexceptions -mfloat-abi=softfp -mfpu=neon -std=c++ -Wno-deprecated -ftree-vectorize -ffast-math -fsingle-precision-constant APP_ABI := arm64-v8a # armeabi-v7a APP_OPTIM := release NDK_TOOLCHAIN_VERSION := clang++ APP_PLATFORM := android-26 |
使用ndk-build编译
|
1 2 |
cd ~/apps/clapack_3.2.1 ndk-build NDK_PROJECT_PATH=. APP_BUILD_SCRIPT=./Android.mk NDK_APPLICATION_MK=./Application.mk |
|
1 2 3 4 5 6 |
[arm64-v8a] Compile : f2c <= xwsne.c [arm64-v8a] Compile : f2c <= dtime_.c [arm64-v8a] Compile : f2c <= etime_.c [arm64-v8a] StaticLibrary : libf2c.a [arm64-v8a] SharedLibrary : liblapack.so [arm64-v8a] Install : liblapack.so => libs/arm64-v8a/liblapack.so |
结束后，复制生成的 ~/apps/clapack_3.2.1/obj/local/armeabi-v8a/ 里的全部文件
libblas.a libclapack.a libf2c.a liblapack.so objs(文件夹)
到 ~/apps/openblas_0.3.18/install/lib 目录。如果没有liblapack.a，就复制 libclapack.a 为 liblapack.a 。
编译时tmglib模块缺失参考（来自 九月是我的猫 Mr-zhu0912）：
4.编译kaldi
下载
|
1 |
git clone https://github.com/kaldi-asr/kaldi ~/apps/kaldi_android |
本例kaldi的commit id：66f5434d29e2a528b9363e0fa25f2793069602a3
使用如下指令检查依赖库和工具是否齐全。
|
1 2 |
cd ~/apps/kaldi_android/tools/ extras/check_dependencies.sh |
根据提示，linux使用ap-get、macos使用brew进行安装。
4.1.编译openfst
本步骤可参考 ~/apps/kaldi_android/tools/Makefile 文件内容，约55行开始。
不要直接在tools目录下执行make编译，否则编译kaldi时可能有下面的openfst错误：
~/apps/kaldi_android/tools/openfst-1.7.2/lib/libfst.a: error adding symbols: File in wrong format
clang80++: error: linker command failed with exit code 1 (use -v to see invocation)
下载地址：https://www.openfst.org/twiki/bin/view/FST/FstDownload
本例为openfst-1.7.2.tar.gz，解压到kaldi目录下 ~/apps/kaldi_android/tools/openfst-1.7.2。
|
1 2 3 4 5 6 7 8 9 10 11 12 |
cd ~/apps/kaldi_android/tools/openfst-1.7.2 ./configure \ --prefix=`pwd` \ --enable-static \ --enable-shared \ --enable-far \ --enable-ngram-fsts \ --host=aarch64-linux-android \ --cache-file=aarch64-linux-android.cache \ --enable-lookahead-fsts \ --with-pic \ LIBS="-ldl" |
如果报错：
configure: error: cannot run test program while cross compiling
编辑 configure 文件，约16776行，注释掉if块内语句：
|
1 2 3 4 5 6 |
if test "$cross_compiling" = yes; then : # { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 #$as_echo "$as_me: error: in \`$ac_pwd':" >&2;} #as_fn_error $? "cannot run test program while cross compiling #See \`config.log' for more details" "$LINENO" 5; } else |
编译：
|
1 2 3 4 |
make -j4 make install cd ../ ln -s openfst-1.7.2 openfst |
4.2.编译kaldi
进入 ~/apps/kaldi_android/src/matrix，打开Makefile 文件，
将 TESTFILES 一行注释。
进入 ~/apps/kaldi_android/src，
打开Makefile，SUBDIRS配置了哪些模块会被编译。可以根据需要删除手机上用不到的模块，也可以添加自己开发的模块。
可以看到本例的kaldi默认只有online2，而没有online。
打开configure文件，
（1）找到约750行，
--android-incdir=*) 一项配置，设置 dynamic_kaldi=true 。
（2）约952行，
if [[ "$use_cuda" = true && ! -f $CUBROOT/cub/cub.cuh ]]; then 代码块，整个注释，关闭英伟达的GPU库cuda。
创建kaldi.mk：
|
1 2 3 4 5 6 7 |
cd ~/apps/kaldi_android/src CXX=clang++ \ ./configure --shared \ --openblas-root=~/apps/openblas_0.3.18/install \ # 《2.编译openblas》的路径 --android-incdir=~/apps/android-ndk-r20b/arm64-toolchain/sysroot/usr/include \ --host=aarch64-linux-android \ # arm-linux-androideabi对应32位，aarch64-linux-android对应64位ARMv8。 --use-cuda=no |
|
1 2 3 4 5 6 7 8 9 10 11 12 13 |
Configuring KALDI to use OPENBLAS. Checking compiler aarch64-linux-android-clang++ ... Checking OpenFst library in ~/apps/kaldi/tools/openfst-1.7.2 ... Performing OS specific configuration ... Using OpenBLAS as the linear algebra library. Successfully configured for Android with OpenBLAS from ~/apps/openblas_0.3.18/install. Kaldi has been successfully configured. To compile: make -j clean depend; make -j <NCPU> where <NCPU> is the number of parallel builds you can afford to do. If unsure, use the smaller of the number of CPUs or the amount of RAM in GB divided by 2, to stay within safe limits. 'make -j' without the numeric value may not limit the number of parallel jobs at all, and overwhelm even a powerful workstation, since Kaldi build is highly parallelized. |
如果提示：
***configure failed: aarch64-linux-android-clang++ is not installed.
You need GNU g++ >= 5.0, Apple clang >= 6.0 or LLVM clang >= 3.5. ***
可能是 --android-incdir 设置的路径未访问到，检查《1.准备toolchain》的toolchain的路径。
在 ~/apps/kaldi_android/src 目录生成kaldi.mk文件。打开修改：
约57行一项，末尾行注释并添加一条 -O3 -DNDEBUG。释义：https://www.cnblogs.com/x0rjchen/p/9310671.html 。
c++版本改为11，默认14.
|
1 2 3 4 5 6 7 8 |
CXXFLAGS = -std=c++11 -I.. -I$(OPENFSTINC) -O1 $(EXTRA_CXXFLAGS) \ -Wall -Wno-sign-compare -Wno-unused-local-typedefs \ -Wno-deprecated-declarations -Winit-self -Wno-mismatched-tags \ -DKALDI_DOUBLEPRECISION=$(DOUBLE_PRECISION) \ -DHAVE_CXXABI_H -DHAVE_OPENBLAS -DANDROID_BUILD \ -I$(OPENBLASINC) -I$(ANDROIDINC) -ftree-vectorize -mfloat-abi=softfp \ -mfpu=neon -pthread \ -O3 -DNDEBUG # -g # -O0 -DKALDI_PARANOID |
编译：
|
1 2 |
make -j4 clean depend make -j4 |
生成的so文件可在 ~/apps/kaldi_android/src/lib 下找到。
这些都是软连接，可以使用cp命令拷贝22个对应的实际so文件到已存在的目录 ~/Desktop/kaldi-lib ：
|
1 2 |
cd ~/apps/kaldi_android/src/lib cp -L ./* ~/Desktop/kaldi-lib |
可以使用aarch64-linux-android-readelf查看so文件的信息。
~/apps/android-ndk-r20b/toolchains/aarch64-linux-android-4.9/prebuilt/darwin-x86_64/bin/aarch64-linux-android-readelf ，
或 ~/apps/android-ndk-r20b/toolchains/llvm/prebuilt/darwin-x86_64/bin/aarch64-linux-android-readelf。
5.cmake集成kaldi
本章节内容谨供参考，实际开发通常只在手机上做识别，用不到这么多头文件和库文件。
本例手机MiNote3，处理器是骁龙660，MIUI 12.0.1（Android 9）。
|
1 2 3 4 5 6 |
使用adb指令可以查看cpu架构。 adb shell getprop ro.product.cpu.abi 获取系统版本： adb shell getprop ro.build.version.release 获取系统api版本 adb shell getprop ro.build.version.sdk |
5.1.准备h文件
从 ~/apps/kaldi_android/src ，提取不包含“bin”结尾的文件夹，并且只提取h文件。
使用下面的指令和python代码，将目标文件拷贝到 ~/Desktop/kaldi_src_h 。
● 第1步shell，获取src下的文件夹名，即模块名。也可直接在python中提取：
|
1 2 3 |
$ ls -l | grep ^d | grep -v "bin" | awk 'BEGIN {FS=" "} {print $9}' | awk 'BEGIN{RS="\n"; ORS="\", \"";}{print $0}' |
● 第2步python：
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
import os import glob from shutil import copyfile fs = ["base", "chain", "cudadecoder", "cudafeat", "cudamatrix", "decoder", "doc", "feat", "fstext", "gmm", "gst-plugin", "hmm", "itf", "ivector", "kws", "lat", "lm", "makefiles", "matrix", "nnet", "nnet2", "nnet3", "online", "online2", "probe", "rnnlm", "sgmm2", "tfrnnlm", "transform", "tree", "util"] src = "~/apps/kaldi_android/src" dst = "~/Desktop/kaldi_src_h" for folder in fs: dst_folder = os.path.join(dst, folder) if not os.path.exists(dst_folder): os.makedirs(dst_folder) src_dir = os.path.join(src, folder) h_dir = os.path.join(src_dir, '*.h') h_list = glob.glob(h_dir) if len(h_list) < 1: continue for h in h_list: name = os.path.basename(h) dist_file = os.path.join(dst_folder, name) copyfile(h, dist_file) print("finish") |
去除cuda开头的，
添加fst的 ~/apps/kaldi/tools/openfst-1.7.2/src/include/fst 。
5.2.准备so文件
《4.2.编译kaldi》步骤已经提取到 ~/Desktop/kaldi-lib 。
5.3.Android项目配置
创建一个普通的安卓项目。
● 在java下像通常一样编写本地类方法KaldiUtil.java。
● 在main下创建cpp文件夹，
▶ 在cpp下创建CMakeLists.txt文件，
▶ 在cpp下放入自己编写的本地方法cpp及自定义的h 。
▶ 在cpp下创建一个kaldi文件夹，
★ 在kaldi下创建head和lib文件夹，
◆ head中放入h文件，
◆ lib中放入so文件。
5.3.1.本地方法java类
KaldiUtil.java ：
|
1 2 3 4 5 6 7 8 |
package com.cuiweiyou.kaldiandroid.util; public class KaldiUtil { static { System.loadLibrary("KaldiUtil"); // 加载cmake生成的库。这个库名对应CMakeLists.txt中的配置 } public native double KaldiMathLogAdd(double x, double y); // 本地方法 } |
右键 这个文件，Open in Terminal，使用javac指令生成对应的h文件，
把这个h文件改名为KaldiUtil.cpp，放到src/main/cpp下。
|
1 |
javac -encoding UTF-8 -classpath ~/apps/android-sdk/platforms/android-26 -h . KaldiUtil.java |
5.3.2.本地方法cpp实现
KaldiUtil.cpp ：
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
// // 这个原本是javac生成的com_cuiweiyou_kaldiandroid_util_KaldiUtil.h文件， // 直接改后缀名为cpp，并修改代码。 // // 此cpp文件在jni入口 CMakeLists.txt 中引入进行编译。 // // 引用Android准备好的 #include <cstring> #include <jni.h> #include <cinttypes> #include <android/log.h> // 这里引用kaldi及自己写的 #include "kaldi.h" // 这里可以直接include自己需要的h。 #define LOG_TAG "System.out" #define LOGE(...) __android_log_print(ANDROID_LOG_ERROR, LOG_TAG, __VA_ARGS__) // 调用Android的log。 #ifndef _Included_com_cuiweiyou_kaldiandroid_util_KaldiUtil #define _Included_com_cuiweiyou_kaldiandroid_util_KaldiUtil #ifdef __cplusplus extern "C" { #endif /* * 这个是原h文件声明的函数，这里直接修改为实现函数。 * Class: com_cuiweiyou_kaldiandroid_util_KaldiUtil * Method: KaldiMathLogAdd * Signature: (DD)D */ JNIEXPORT jdouble JNICALL Java_com_cuiweiyou_kaldiandroid_util_KaldiUtil_KaldiMathLogAdd (JNIEnv *jniEnv, jobject, jdouble x, jdouble y){ LOGE("进入JNI的范畴了", ""); LOGE("调用kaldi的函数", ""); double v = kaldi::LogAdd(x, y); // 调用 base/kaldi-math.h 中的函数 LOGE("kaldi计算结束，返回java", ""); return v; }; #ifdef __cplusplus } #endif #endif |
kaldi.h不是必须的，这里仅供演示：
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 |
// // 统一说明及管理 // // #ifndef KALDIANDROID_KALDI_H #define KALDIANDROID_KALDI_H // kaldi基础类 // #include "base/io-funcs-inl.h" // #include "base/io-funcs.h" // #include "base/kaldi-common.h" // #include "base/kaldi-error.h" #include "base/kaldi-math.h" // ------------------------------------- 本例使用的h // #include "base/kaldi-types.h" // #include "base/kaldi-utils.h" // #include "base/timer.h" // #include "base/version.h" // Chain模型基础类 // #include "chain/chain-datastruct.h" // #include "chain/chain-den-graph.h" // #include "chain/chain-denominator.h" // #include "chain/chain-generic-numerator.h" // #include "chain/chain-kernels-ansi.h" // #include "chain/chain-numerator.h" // #include "chain/chain-supervision.h" // #include "chain/chain-training.h" // #include "chain/language-model.h" // cuda开头的是英伟达显卡相关api。目前手机都无法使用 // #include "cudadecoder/batched-static-nnet3-kernels.h" // #include "cudadecoder/batched-static-nnet3.h" // #include "cudadecoder/batched-threaded-nnet3-cuda-online-pipeline.h" // #include "cudadecoder/batched-threaded-nnet3-cuda-pipeline.h" // #include "cudadecoder/batched-threaded-nnet3-cuda-pipeline2.h" // #include "cudadecoder/cuda-decodable-itf.h" // #include "cudadecoder/cuda-decoder-common.h" // #include "cudadecoder/cuda-decoder-kernels-utils.h" // #include "cudadecoder/cuda-decoder-kernels.h" // #include "cudadecoder/cuda-decoder.h" // #include "cudadecoder/cuda-fst.h" // #include "cudadecoder/cuda-online-pipeline-dynamic-batcher.h" // #include "cudadecoder/cuda-pipeline-common.h" // #include "cudadecoder/decodable-cumatrix.h" // #include "cudadecoder/lattice-postprocessor.h" // #include "cudadecoder/thread-pool-light.h" // #include "cudadecoder/thread-pool.h" // -- // #include "cudafeat/feature-online-batched-cmvn-cuda-kernels.h" // #include "cudafeat/feature-online-batched-cmvn-cuda.h" // #include "cudafeat/feature-online-batched-ivector-cuda-kernels.h" // #include "cudafeat/feature-online-batched-ivector-cuda.h" // #include "cudafeat/feature-online-batched-spectral-cuda-kernels.h" // #include "cudafeat/feature-online-batched-spectral-cuda.h" // #include "cudafeat/feature-online-cmvn-cuda.h" // #include "cudafeat/feature-spectral-cuda.h" // #include "cudafeat/feature-window-cuda.h" // #include "cudafeat/lane-desc.h" // #include "cudafeat/online-batched-feature-pipeline-cuda.h" // #include "cudafeat/online-cuda-feature-pipeline.h" // #include "cudafeat/online-ivector-feature-cuda-kernels.h" // #include "cudafeat/online-ivector-feature-cuda.h" // -- // #include "cudamatrix/cu-allocator.h" // #include "cudamatrix/cu-array-inl.h" // #include "cudamatrix/cu-array.h" // #include "cudamatrix/cu-block-matrix.h" // #include "cudamatrix/cu-common.h" // #include "cudamatrix/cu-compressed-matrix.h" // #include "cudamatrix/cu-device.h" // #include "cudamatrix/cu-kernels-ansi.h" // #include "cudamatrix/cu-kernels.h" // #include "cudamatrix/cu-math.h" // #include "cudamatrix/cu-matrix-inl.h" // #include "cudamatrix/cu-matrix-lib.h" // #include "cudamatrix/cu-matrix.h" // #include "cudamatrix/cu-matrixdim.h" // #include "cudamatrix/cu-packed-matrix.h" // #include "cudamatrix/cu-rand.h" // #include "cudamatrix/cu-sp-matrix.h" // #include "cudamatrix/cu-sparse-matrix.h" // #include "cudamatrix/cu-tp-matrix.h" // #include "cudamatrix/cu-value.h" // #include "cudamatrix/cu-vector.h" // #include "cudamatrix/cublas-wrappers.h" // 解码器相关 // #include "decoder/biglm-faster-decoder.h" // #include "decoder/decodable-mapped.h" // #include "decoder/decodable-matrix.h" // #include "decoder/decodable-sum.h" // #include "decoder/decoder-wrappers.h" // #include "decoder/faster-decoder.h" // #include "decoder/grammar-fst.h" // #include "decoder/lattice-biglm-faster-decoder.h" // #include "decoder/lattice-faster-decoder.h" // #include "decoder/lattice-faster-online-decoder.h" // #include "decoder/lattice-incremental-decoder.h" // #include "decoder/lattice-incremental-online-decoder.h" // #include "decoder/lattice-simple-decoder.h" // #include "decoder/simple-decoder.h" // #include "decoder/training-graph-compiler.h" // 特征提取相关。手机不做相关工作，用不到 // #include "feat/feature-common-inl.h" // #include "feat/feature-common.h" // #include "feat/feature-fbank.h" // #include "feat/feature-functions.h" // #include "feat/feature-mfcc.h" // #include "feat/feature-plp.h" // #include "feat/feature-spectrogram.h" // #include "feat/feature-window.h" // #include "feat/mel-computations.h" // #include "feat/online-feature.h" // #include "feat/pitch-functions.h" // #include "feat/resample.h" // #include "feat/signal.h" // #include "feat/wave-reader.h" // OpenFST有限自动机 相关 // #include "fst/accumulator.h" // #include "fst/add-on.h" // #include "fst/arc-arena.h" // #include "fst/arc-map.h" // #include "fst/arc.h" // #include "fst/arcfilter.h" // #include "fst/arcsort.h" // #include "fst/bi-table.h" // #include "fst/cache.h" // #include "fst/closure.h" // #include "fst/compact-fst.h" // #include "fst/compat.h" // #include "fst/complement.h" // #include "fst/compose-filter.h" // #include "fst/compose.h" // #include "fst/concat.h" // #include "fst/config.h" // #include "fst/connect.h" // #include "fst/const-fst.h" // #include "fst/determinize.h" // #include "fst/dfs-visit.h" // #include "fst/difference.h" // #include "fst/disambiguate.h" // #include "fst/edit-fst.h" // #include "fst/encode.h" // #include "fst/epsnormalize.h" // #include "fst/equal.h" // #include "fst/equivalent.h" // #include "fst/expanded-fst.h" // #include "fst/expectation-weight.h" // #include "fst/factor-weight.h" // #include "fst/filter-state.h" // #include "fst/flags.h" // #include "fst/float-weight.h" // #include "fst/fst-decl.h" // #include "fst/fst.h" // #include "fst/fstlib.h" // #include "fst/generic-register.h" // #include "fst/heap.h" // #include "fst/icu.h" // #include "fst/intersect.h" // #include "fst/interval-set.h" // #include "fst/invert.h" // #include "fst/isomorphic.h" // #include "fst/label-reachable.h" // #include "fst/lexicographic-weight.h" // #include "fst/lock.h" // #include "fst/log.h" // #include "fst/lookahead-filter.h" // #include "fst/lookahead-matcher.h" // #include "fst/map.h" // #include "fst/mapped-file.h" // #include "fst/matcher-fst.h" // #include "fst/matcher.h" // #include "fst/memory.h" // #include "fst/minimize.h" // #include "fst/mutable-fst.h" // #include "fst/pair-weight.h" // #include "fst/partition.h" // #include "fst/power-weight.h" // #include "fst/product-weight.h" // #include "fst/project.h" // #include "fst/properties.h" // #include "fst/prune.h" // #include "fst/push.h" // #include "fst/queue.h" // #include "fst/randequivalent.h" // #include "fst/randgen.h" // #include "fst/rational.h" // #include "fst/register.h" // #include "fst/relabel.h" // #include "fst/replace-util.h" // #include "fst/replace.h" // #include "fst/reverse.h" // #include "fst/reweight.h" // #include "fst/rmepsilon.h" // #include "fst/rmfinalepsilon.h" // #include "fst/set-weight.h" // #include "fst/shortest-distance.h" // #include "fst/shortest-path.h" // #include "fst/signed-log-weight.h" // #include "fst/sparse-power-weight.h" // #include "fst/sparse-tuple-weight.h" // #include "fst/state-map.h" // #include "fst/state-reachable.h" // #include "fst/state-table.h" // #include "fst/statesort.h" // #include "fst/string-weight.h" // #include "fst/string.h" // #include "fst/symbol-table-ops.h" // #include "fst/symbol-table.h" // #include "fst/synchronize.h" // #include "fst/test-properties.h" // #include "fst/topsort.h" // #include "fst/tuple-weight.h" // #include "fst/types.h" // #include "fst/union-find.h" // #include "fst/union-weight.h" // #include "fst/union.h" // #include "fst/util.h" // #include "fst/vector-fst.h" // #include "fst/verify.h" // #include "fst/visit.h" // #include "fst/weight.h" // --openfst扩展 // #include "fstext/context-fst.h" // #include "fstext/deterministic-fst-inl.h" // #include "fstext/deterministic-fst.h" // #include "fstext/determinize-lattice-inl.h" // #include "fstext/determinize-lattice.h" // #include "fstext/determinize-star-inl.h" // #include "fstext/determinize-star.h" // #include "fstext/epsilon-property-inl.h" // #include "fstext/epsilon-property.h" // #include "fstext/factor-inl.h" // #include "fstext/factor.h" // #include "fstext/fst-test-utils.h" // #include "fstext/fstext-lib.h" // #include "fstext/fstext-utils-inl.h" // #include "fstext/fstext-utils.h" // #include "fstext/grammar-context-fst.h" // #include "fstext/kaldi-fst-io-inl.h" // #include "fstext/kaldi-fst-io.h" // #include "fstext/lattice-utils-inl.h" // #include "fstext/lattice-utils.h" // #include "fstext/lattice-weight.h" // #include "fstext/pre-determinize-inl.h" // #include "fstext/pre-determinize.h" // #include "fstext/prune-special-inl.h" // #include "fstext/prune-special.h" // #include "fstext/push-special.h" // #include "fstext/rand-fst.h" // #include "fstext/remove-eps-local-inl.h" // #include "fstext/remove-eps-local.h" // #include "fstext/table-matcher.h" // #include "fstext/trivial-factor-weight.h" // GMM (对角阵高斯混合) 模型基础类 // #include "gmm/am-diag-gmm.h" // #include "gmm/decodable-am-diag-gmm.h" // #include "gmm/diag-gmm-inl.h" // #include "gmm/diag-gmm-normal.h" // #include "gmm/diag-gmm.h" // #include "gmm/ebw-diag-gmm.h" // #include "gmm/full-gmm-inl.h" // #include "gmm/full-gmm-normal.h" // #include "gmm/full-gmm.h" // #include "gmm/indirect-diff-diag-gmm.h" // #include "gmm/mle-am-diag-gmm.h" // #include "gmm/mle-diag-gmm.h" // #include "gmm/mle-full-gmm.h" // #include "gmm/model-common.h" // #include "gmm/model-test-common.h" // GStreamer解码器插件 // #include "gst-plugin/gst-audio-source.h" // #include "gst-plugin/gst-online-gmm-decode-faster.h" // HMM模型相关 // #include "hmm/hmm-test-utils.h" // #include "hmm/hmm-topology.h" // #include "hmm/hmm-utils.h" // #include "hmm/posterior.h" // #include "hmm/transition-model.h" // #include "hmm/tree-accu.h" // 扩展接口 // #include "itf/clusterable-itf.h" // #include "itf/context-dep-itf.h" // #include "itf/decodable-itf.h" // #include "itf/online-feature-itf.h" // #include "itf/optimizable-itf.h" // #include "itf/options-itf.h" // #include "itf/transition-information.h" // 声纹识别(说话人识别) // #include "ivector/agglomerative-clustering.h" // #include "ivector/ivector-extractor.h" // #include "ivector/logistic-regression.h" // #include "ivector/plda.h" // #include "ivector/voice-activity-detection.h" // Keyword Search关键词搜索 // #include "kws/kaldi-kws.h" // #include "kws/kws-functions.h" // #include "kws/kws-scoring.h" // 构建词图 // #include "lat/arctic-weight.h" // #include "lat/compose-lattice-pruned.h" // #include "lat/confidence.h" // #include "lat/determinize-lattice-pruned.h" // #include "lat/kaldi-lattice.h" // #include "lat/lattice-functions-transition-model.h" // #include "lat/lattice-functions.h" // #include "lat/minimize-lattice.h" // #include "lat/phone-align-lattice.h" // #include "lat/push-lattice.h" // #include "lat/sausages.h" // #include "lat/word-align-lattice-lexicon.h" // #include "lat/word-align-lattice.h" // 自带的language model语言模型 // #include "lm/arpa-file-parser.h" // #include "lm/arpa-lm-compiler.h" // #include "lm/const-arpa-lm.h" // #include "lm/kaldi-rnnlm.h" // #include "lm/kenlm.h" // #include "lm/mikolov-rnnlm-lib.h" // 矩阵计算 // #include "matrix/cblas-wrappers.h" // #include "matrix/compressed-matrix.h" // #include "matrix/jama-eig.h" // #include "matrix/jama-svd.h" // #include "matrix/kaldi-blas.h" // #include "matrix/kaldi-matrix-inl.h" // #include "matrix/kaldi-matrix.h" // #include "matrix/kaldi-vector-inl.h" // #include "matrix/kaldi-vector.h" // #include "matrix/matrix-common.h" // #include "matrix/matrix-functions-inl.h" // #include "matrix/matrix-functions.h" // #include "matrix/matrix-lib.h" // #include "matrix/numpy-array.h" // #include "matrix/optimization.h" // #include "matrix/packed-matrix.h" // #include "matrix/sp-matrix-inl.h" // #include "matrix/sp-matrix.h" // #include "matrix/sparse-matrix.h" // #include "matrix/srfft.h" // #include "matrix/tp-matrix.h" // nnetX，训练方面的工作手机干不动。 // DNN训练声学模型实现方式1，仅支持单GPU，维护者Karel // #include "nnet/nnet-activation.h" // #include "nnet/nnet-affine-transform.h" // #include "nnet/nnet-average-pooling-component.h" // #include "nnet/nnet-blstm-projected.h" // #include "nnet/nnet-component.h" // #include "nnet/nnet-convolutional-component.h" // #include "nnet/nnet-frame-pooling-component.h" // #include "nnet/nnet-kl-hmm.h" // #include "nnet/nnet-linear-transform.h" // #include "nnet/nnet-loss.h" // #include "nnet/nnet-lstm-projected.h" // #include "nnet/nnet-matrix-buffer.h" // #include "nnet/nnet-max-pooling-component.h" // #include "nnet/nnet-multibasis-component.h" // #include "nnet/nnet-nnet.h" // #include "nnet/nnet-parallel-component.h" // #include "nnet/nnet-parametric-relu.h" // #include "nnet/nnet-pdf-prior.h" // #include "nnet/nnet-randomizer.h" // #include "nnet/nnet-rbm.h" // #include "nnet/nnet-recurrent.h" // #include "nnet/nnet-sentence-averaging-component.h" // #include "nnet/nnet-trnopts.h" // #include "nnet/nnet-utils.h" // #include "nnet/nnet-various.h" // DNN训练声学模型实现方式2，支持多GPU、CPU，维护者Daniel // #include "nnet2/am-nnet.h" // #include "nnet2/combine-nnet-a.h" // #include "nnet2/combine-nnet-fast.h" // #include "nnet2/combine-nnet.h" // #include "nnet2/decodable-am-nnet.h" // #include "nnet2/get-feature-transform.h" // #include "nnet2/mixup-nnet.h" // #include "nnet2/nnet-component.h" // #include "nnet2/nnet-compute-discriminative-parallel.h" // #include "nnet2/nnet-compute-discriminative.h" // #include "nnet2/nnet-compute-online.h" // #include "nnet2/nnet-compute.h" // #include "nnet2/nnet-example-functions.h" // #include "nnet2/nnet-example.h" // #include "nnet2/nnet-fix.h" // #include "nnet2/nnet-functions.h" // #include "nnet2/nnet-limit-rank.h" // #include "nnet2/nnet-nnet.h" // #include "nnet2/nnet-precondition-online.h" // #include "nnet2/nnet-precondition.h" // #include "nnet2/nnet-stats.h" // #include "nnet2/nnet-update-parallel.h" // #include "nnet2/nnet-update.h" // #include "nnet2/online-nnet2-decodable.h" // #include "nnet2/rescale-nnet.h" // #include "nnet2/shrink-nnet.h" // #include "nnet2/train-nnet-ensemble.h" // #include "nnet2/train-nnet.h" // #include "nnet2/widen-nnet.h" // DNN训练声学模型实现方式3，nnet2的改进，维护者Daniel // #include "nnet3/am-nnet-simple.h" // #include "nnet3/attention.h" // #include "nnet3/convolution.h" // #include "nnet3/decodable-batch-looped.h" // #include "nnet3/decodable-online-looped.h" // #include "nnet3/decodable-simple-looped.h" // #include "nnet3/discriminative-supervision.h" // #include "nnet3/discriminative-training.h" // #include "nnet3/natural-gradient-online.h" // #include "nnet3/nnet-am-decodable-simple.h" // #include "nnet3/nnet-analyze.h" // #include "nnet3/nnet-attention-component.h" // #include "nnet3/nnet-batch-compute.h" // #include "nnet3/nnet-chain-diagnostics.h" // #include "nnet3/nnet-chain-diagnostics2.h" // #include "nnet3/nnet-chain-example.h" // #include "nnet3/nnet-chain-training.h" // #include "nnet3/nnet-chain-training2.h" // #include "nnet3/nnet-combined-component.h" // #include "nnet3/nnet-common.h" // #include "nnet3/nnet-compile-looped.h" // #include "nnet3/nnet-compile-utils.h" // #include "nnet3/nnet-compile.h" // #include "nnet3/nnet-component-itf.h" // #include "nnet3/nnet-computation-graph.h" // #include "nnet3/nnet-computation.h" // #include "nnet3/nnet-compute.h" // #include "nnet3/nnet-convolutional-component.h" // #include "nnet3/nnet-descriptor.h" // #include "nnet3/nnet-diagnostics.h" // #include "nnet3/nnet-discriminative-diagnostics.h" // #include "nnet3/nnet-discriminative-example.h" // #include "nnet3/nnet-discriminative-training.h" // #include "nnet3/nnet-example-utils.h" // #include "nnet3/nnet-example.h" // #include "nnet3/nnet-general-component.h" // #include "nnet3/nnet-graph.h" // #include "nnet3/nnet-nnet.h" // #include "nnet3/nnet-normalize-component.h" // #include "nnet3/nnet-optimize-utils.h" // #include "nnet3/nnet-optimize.h" // #include "nnet3/nnet-parse.h" // #include "nnet3/nnet-simple-component.h" // #include "nnet3/nnet-test-utils.h" // #include "nnet3/nnet-training.h" // #include "nnet3/nnet-utils.h" // online是用来识别的，手机上用到。 // 语音识别第1版api，停更了。 // #include "online/online-audio-source.h" // #include "online/online-decodable.h" // #include "online/online-faster-decoder.h" // #include "online/online-feat-input.h" // #include "online/online-tcp-source.h" // #include "online/onlinebin-util.h" // 语音识别第2版，较复杂 // #include "online2/online-endpoint.h" // #include "online2/online-feature-pipeline.h" // #include "online2/online-gmm-decodable.h" // #include "online2/online-gmm-decoding.h" // #include "online2/online-ivector-feature.h" // #include "online2/online-nnet2-decoding-threaded.h" // #include "online2/online-nnet2-decoding.h" // #include "online2/online-nnet2-feature-pipeline.h" // #include "online2/online-nnet3-decoding.h" // #include "online2/online-nnet3-incremental-decoding.h" // #include "online2/online-nnet3-wake-word-faster-decoder.h" // #include "online2/online-speex-wrapper.h" // #include "online2/online-timing.h" // #include "online2/onlinebin-util.h" // RNN语言模型训练。手机干不动 // #include "rnnlm/rnnlm-compute-state.h" // #include "rnnlm/rnnlm-core-compute.h" // #include "rnnlm/rnnlm-core-training.h" // #include "rnnlm/rnnlm-embedding-training.h" // #include "rnnlm/rnnlm-example-utils.h" // #include "rnnlm/rnnlm-example.h" // #include "rnnlm/rnnlm-lattice-rescoring.h" // #include "rnnlm/rnnlm-test-utils.h" // #include "rnnlm/rnnlm-training.h" // #include "rnnlm/rnnlm-utils.h" // #include "rnnlm/sampler.h" // #include "rnnlm/sampling-lm-estimate.h" // #include "rnnlm/sampling-lm.h" // SGMM模型训练。手机干不动 // #include "sgmm2/am-sgmm2-project.h" // #include "sgmm2/am-sgmm2.h" // #include "sgmm2/decodable-am-sgmm2.h" // #include "sgmm2/estimate-am-sgmm2-ebw.h" // #include "sgmm2/estimate-am-sgmm2.h" // #include "sgmm2/fmllr-sgmm2.h" // 基于tensorflow RNN语言模型训练。手机干不动 // #include "tfrnnlm/tensorflow-rnnlm.h" // 特征转换相关 // #include "transform/basis-fmllr-diag-gmm.h" // #include "transform/cmvn.h" // #include "transform/compressed-transform-stats.h" // #include "transform/decodable-am-diag-gmm-regtree.h" // #include "transform/fmllr-diag-gmm.h" // #include "transform/fmllr-raw.h" // #include "transform/fmpe.h" // #include "transform/lda-estimate.h" // #include "transform/lvtln.h" // #include "transform/mllt.h" // #include "transform/regression-tree.h" // #include "transform/regtree-fmllr-diag-gmm.h" // #include "transform/regtree-mllr-diag-gmm.h" // #include "transform/transform-common.h" // 内部决策树相关 // #include "tree/build-tree-questions.h" // #include "tree/build-tree-utils.h" // #include "tree/build-tree.h" // #include "tree/cluster-utils.h" // #include "tree/clusterable-classes.h" // #include "tree/context-dep.h" // #include "tree/event-map.h" // #include "tree/tree-renderer.h" // 基础工具类 // #include "util/basic-filebuf.h" // #include "util/common-utils.h" // #include "util/const-integer-set-inl.h" // #include "util/const-integer-set.h" // #include "util/edit-distance-inl.h" // #include "util/edit-distance.h" // #include "util/hash-list-inl.h" // #include "util/hash-list.h" // #include "util/kaldi-cygwin-io-inl.h" // #include "util/kaldi-holder-inl.h" // #include "util/kaldi-holder.h" // #include "util/kaldi-io-inl.h" // #include "util/kaldi-io.h" // #include "util/kaldi-pipebuf.h" // #include "util/kaldi-semaphore.h" // #include "util/kaldi-table-inl.h" // #include "util/kaldi-table.h" // #include "util/kaldi-thread.h" // #include "util/parse-options.h" // #include "util/simple-io-funcs.h" // #include "util/simple-options.h" // #include "util/stl-utils.h" // #include "util/table-types.h" // #include "util/text-utils.h" #endif //KALDIANDROID_KALDI_H |
5.3.3.CMakeLists.txt
这个文件的写法很多，官方教程 https://cmake.org/cmake/help/latest/guide/tutorial/index.html 。
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
cmake_minimum_required(VERSION 3.4.1) # ${CMAKE_SOURCE_DIR} 即本CMakeLists.txt文件所在目录 # 设置一个路径变量KALDI_DIR，对应与CMakeLists.txt同级的 ./kaldi set(KALDI_DIR ${CMAKE_SOURCE_DIR}/kaldi) # ${ANDROID_ABI} 对应 build.gradle文件 android / defaultConfig / ndk / abiFilters 的芯片架构 add_library(libkaldi-base SHARED IMPORTED) set_target_properties(libkaldi-base PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-base.so) add_library(libkaldi-chain SHARED IMPORTED) set_target_properties(libkaldi-chain PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-chain.so) add_library(libkaldi-decoder SHARED IMPORTED) set_target_properties(libkaldi-decoder PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-decoder.so) add_library(libkaldi-feat SHARED IMPORTED) set_target_properties(libkaldi-feat PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-feat.so) add_library(libkaldi-fstext SHARED IMPORTED) set_target_properties(libkaldi-fstext PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-fstext.so) add_library(libkaldi-gmm SHARED IMPORTED) set_target_properties(libkaldi-gmm PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-gmm.so) add_library(libkaldi-hmm SHARED IMPORTED) set_target_properties(libkaldi-hmm PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-hmm.so) add_library(libkaldi-ivector SHARED IMPORTED) set_target_properties(libkaldi-ivector PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-ivector.so) add_library(libkaldi-kws SHARED IMPORTED) set_target_properties(libkaldi-kws PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-kws.so) add_library(libkaldi-lat SHARED IMPORTED) set_target_properties(libkaldi-lat PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-lat.so) add_library(libkaldi-lm SHARED IMPORTED) set_target_properties(libkaldi-lm PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-lm.so) add_library(libkaldi-matrix SHARED IMPORTED) set_target_properties(libkaldi-matrix PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-matrix.so) add_library(libkaldi-nnet SHARED IMPORTED) set_target_properties(libkaldi-nnet PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-nnet.so) add_library(libkaldi-nnet2 SHARED IMPORTED) set_target_properties(libkaldi-nnet2 PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-nnet2.so) add_library(libkaldi-nnet3 SHARED IMPORTED) set_target_properties(libkaldi-nnet3 PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-nnet3.so) add_library(libkaldi-online2 SHARED IMPORTED) set_target_properties(libkaldi-online2 PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-online2.so) add_library(libkaldi-rnnlm SHARED IMPORTED) set_target_properties(libkaldi-rnnlm PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-rnnlm.so) add_library(libkaldi-sgmm2 SHARED IMPORTED) set_target_properties(libkaldi-sgmm2 PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-sgmm2.so) add_library(libkaldi-transform SHARED IMPORTED) set_target_properties(libkaldi-transform PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-transform.so) add_library(libkaldi-tree SHARED IMPORTED) set_target_properties(libkaldi-tree PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-tree.so) add_library(libkaldi-util SHARED IMPORTED) set_target_properties(libkaldi-util PROPERTIES IMPORTED_LOCATION ${KALDI_DIR}/lib/${ANDROID_ABI}/libkaldi-util.so) # 自己编写的本地方法类 add_library(KaldiUtil SHARED KaldiUtil.cpp) # 相对CMakeLists.txt的路径 target_include_directories(KaldiUtil PRIVATE ${KALDI_DIR}/head) # 【src/main/cpp/kaldi/head】，自定义本地方法中 include头文件 拼接的路径 # build application's shared lib set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11") # 最终编译的so文件路径：项目/app/build/intermediates/merged_native_libs/debug/out/lib/arm64-v8a target_link_libraries(KaldiUtil android libkaldi-base # 对应本例演示用的 base/kaldi-math.h # libkaldi-chain # libkaldi-decoder # libkaldi-feat # libkaldi-fstext # libkaldi-gmm # libkaldi-hmm # libkaldi-ivector # libkaldi-kws # libkaldi-lat # libkaldi-lm # libkaldi-matrix # libkaldi-nnet # libkaldi-nnet2 # libkaldi-nnet3 # libkaldi-online2 # libkaldi-rnnlm # libkaldi-sgmm2 # libkaldi-transform # libkaldi-tree # libkaldi-util log) |
5.3.4.build.gradle
添加cmake相关项。
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
plugins { id 'com.android.application' } android { compileSdkVersion 30 buildToolsVersion "30.0.3" defaultConfig { applicationId "com.cuiweiyou.kaldiandroid" minSdkVersion 26 // 【对应编译so时的api版本】 targetSdkVersion 30 versionCode 1 versionName "1.0" externalNativeBuild { cmake { arguments '-DANDROID_STL=c++_shared' // 【解决 UnsatisfiedLinkError: dlopen failed: library "libc++_shared.so" not found】 } } ndk { abiFilters "arm64-v8a" // 【对应本手机的芯片，及so的架构】 } } buildTypes { release { minifyEnabled false proguardFiles getDefaultProguardFile('proguard-android-optimize.txt'), 'proguard-rules.pro' } } compileOptions { sourceCompatibility JavaVersion.VERSION_1_8 targetCompatibility JavaVersion.VERSION_1_8 } externalNativeBuild { cmake { path 'src/main/cpp/CMakeLists.txt' // 【JNI入口】 } } } dependencies { // 略 } |
5.3.5.运行测试
进行前几步骤时，cpp代码通常会报红，可以时不时的点击菜单栏Build下的“Make Project”。
在Activity中调用一下：
|
1 2 3 4 5 6 7 8 9 10 11 |
public class MainActivity extends AppCompatActivity { @Override protected void onCreate(Bundle savedInstanceState) { super.onCreate(savedInstanceState); setContentView(R.layout.activity_main); double v = new KaldiUtil().KaldiMathLogAdd(1, 2); // 调用native方法 Log.e("ard", "这里是java。kaldi计算的结果：" + v); } } |
日志：
整个Android项目可以从git下载测试：
https://github.com/731942771/kaldi-android
https://gitee.com/vigiles/kaldi-android
- end
声明
本文由崔维友 威格灵 cuiweiyou vigiles cuiweiyou 原创，转载请注明出处：http://www.gaohaiyan.com/2477.html
承接App定制、企业web站点、办公系统软件 设计开发，外包项目，毕设
