Add VOICEVOX (AI speech synthesis) package for ROS Interface

3rdparty/voicevox/.gitignore

@@ -0,0 +1,6 @@
+build
+dict
+lib
+node_scripts/voicevox_engine
+requirements.txt
+!.gitignore

3rdparty/voicevox/CMakeLists.txt

@@ -0,0 +1,73 @@
+cmake_minimum_required(VERSION 2.8.3)
+project(voicevox)
+
+find_package(catkin REQUIRED
+  COMPONENTS
+  catkin_virtualenv
+)
+
+catkin_python_setup()
+
+set(INSTALL_DIR ${PROJECT_SOURCE_DIR})
+
+catkin_package()
+
+catkin_generate_virtualenv(
+  INPUT_REQUIREMENTS requirements.in
+  PYTHON_INTERPRETER python3
+  USE_SYSTEM_PACKAGES FALSE
+)
+
+add_custom_command(
+  OUTPUT voicevox_model_installed
+  COMMAND make -f ${PROJECT_SOURCE_DIR}/Makefile.model
+  MD5SUM_DIR=${PROJECT_SOURCE_DIR}/md5sum
+  INSTALL_DIR=${INSTALL_DIR}
+)
+
+
+add_custom_command(
+  OUTPUT voicevox_core_installed
+  COMMAND make -f ${PROJECT_SOURCE_DIR}/Makefile.core
+  MD5SUM_DIR=${PROJECT_SOURCE_DIR}/md5sum
+  INSTALL_DIR=${INSTALL_DIR}
+)
+
+add_custom_command(
+  OUTPUT voicevox_engine_installed
+  COMMAND make -f ${PROJECT_SOURCE_DIR}/Makefile.engine
+  MD5SUM_DIR=${PROJECT_SOURCE_DIR}/md5sum
+  INSTALL_DIR=${INSTALL_DIR}
+)
+
+add_custom_command(
+  OUTPUT open_jtalk_dic_installed
+  COMMAND make -f ${PROJECT_SOURCE_DIR}/Makefile.open_jtalk_dic
+  MD5SUM_DIR=${PROJECT_SOURCE_DIR}/md5sum
+  INSTALL_DIR=${INSTALL_DIR}
+)
+
+add_custom_target(all_installed ALL DEPENDS
+  voicevox_model_installed
+  voicevox_core_installed
+  voicevox_engine_installed
+  open_jtalk_dic_installed)
+
+file(GLOB NODE_SCRIPTS_FILES node_scripts/*.py)
+catkin_install_python(
+  PROGRAMS ${NODE_SCRIPTS_FILES}
+  DESTINATION ${CATKIN_PACKAGE_BIN_DESTINATION}/node_scripts/
+)
+install(DIRECTORY node_scripts/voicevox_engine
+  DESTINATION ${CATKIN_PACKAGE_SHARE_DESTINATION}/catkin_virtualenv_scripts/
+  USE_SOURCE_PERMISSIONS)
+install(DIRECTORY launch dict
+  DESTINATION ${CATKIN_PACKAGE_SHARE_DESTINATION}
+  USE_SOURCE_PERMISSIONS)
+install(PROGRAMS bin/text2wave
+  DESTINATION ${CATKIN_PACKAGE_SHARE_DESTINATION}/bin)
+
+install(DIRECTORY
+  ${INSTALL_DIR}/lib
+  DESTINATION ${CATKIN_PACKAGE_SHARE_DESTINATION}
+  USE_SOURCE_PERMISSIONS)

3rdparty/voicevox/Makefile

@@ -0,0 +1,11 @@
+all:
+	make -f Makefile.core
+	make -f Makefile.model
+	make -f Makefile.engine
+	make -f Makefile.open_jtalk_dic
+clean:
+	make -f Makefile.core clean
+	make -f Makefile.model clean
+	make -f Makefile.engine clean
+	make -f Makefile.open_jtalk_dic clean
+	rm -rf build

3rdparty/voicevox/Makefile.core

@@ -0,0 +1,36 @@
+# -*- makefile -*-
+
+all: installed.viocevox_core
+
+VERSION = 0.12.4
+MD5SUM_DIR = $(CURDIR)/md5sum
+ifeq ($(shell uname -m),x86_64)
+	DIRNAME = voicevox_core-linux-x64-cpu-$(VERSION)
+	FILENAME = voicevox_core-linux-x64-cpu-$(VERSION).zip
+	SOURCE_DIR = build/voicevox_core-linux-x64-cpu-$(VERSION)
+	MD5SUM_FILE = $(MD5SUM_DIR)/core-linux-x64-cpu.md5sum
+endif
+ifeq ($(shell uname -m),aarch64)
+	DIRNAME = voicevox_core-linux-arm64-cpu-$(VERSION)
+	FILENAME = voicevox_core-linux-arm64-cpu-$(VERSION).zip
+	SOURCE_DIR = build/voicevox_core-linux-arm64-cpu-$(VERSION)
+	MD5SUM_FILE = $(MD5SUM_DIR)/core-linux-arm64-cpu.md5sum
+endif
+
+TARBALL_URL = "https://github.com/VOICEVOX/voicevox_core/releases/download/$(VERSION)/$(FILENAME)"
+TARBALL = build/$(FILENAME)
+UNPACK_CMD = unzip
+SCRIPT_DIR = $( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
+include $(shell rospack find mk)/download_unpack_build.mk
+INSTALL_DIR =	'./'
+
+
+installed.viocevox_core: $(SOURCE_DIR)/unpacked
+	mkdir -p $(INSTALL_DIR)/lib
+	cp build/$(DIRNAME)/lib*.so $(INSTALL_DIR)/lib/
+
+clean:
+	rm -rf $(TARBALL)
+	rm -rf $(SOURCE_DIR)
+	rm -rf $(INSTALL_DIR)/lib
+	rm -rf build

3rdparty/voicevox/Makefile.engine

@@ -0,0 +1,24 @@
+# -*- makefile -*-
+
+all: installed.voicevox_engine
+
+VERSION = 0.12.4
+FILENAME = $(VERSION).tar.gz
+TARBALL = build/$(FILENAME)
+TARBALL_URL = "https://github.com/VOICEVOX/voicevox_engine/archive/refs/tags/$(FILENAME)"
+SOURCE_DIR = build/voicevox_engine-$(VERSION)
+UNPACK_CMD = tar xvzf
+MD5SUM_DIR = $(CURDIR)/md5sum
+MD5SUM_FILE = $(MD5SUM_DIR)/voicevox_engine.tar.gz.md5sum
+include $(shell rospack find mk)/download_unpack_build.mk
+INSTALL_DIR =	'./'
+
+
+installed.voicevox_engine: $(SOURCE_DIR)/unpacked
+	cp -r build/voicevox_engine-$(VERSION) $(INSTALL_DIR)/node_scripts/voicevox_engine
+
+clean:
+	rm -rf $(TARBALL)
+	rm -rf $(SOURCE_DIR)
+	rm -rf $(INSTALL_DIR)/node_scripts/voicevox_engine
+	rm -rf build

3rdparty/voicevox/Makefile.model

@@ -0,0 +1,33 @@
+# -*- makefile -*-
+
+all: installed.voicevox_model
+
+VERSION = 1.10.0
+ifeq ($(shell uname -m),x86_64)
+	DIRNAME = onnxruntime-linux-x64-$(VERSION)
+endif
+ifeq ($(shell uname -m),aarch64)
+	DIRNAME = onnxruntime-linux-aarch64-$(VERSION)
+endif
+
+SOURCE_DIR = build/$(DIRNAME)
+FILENAME = $(DIRNAME).tgz
+TARBALL = build/$(FILENAME)
+TARBALL_URL = "https://github.com/microsoft/onnxruntime/releases/download/v$(VERSION)/$(FILENAME)"
+UNPACK_CMD = tar xvzf
+MD5SUM_DIR = $(CURDIR)/md5sum
+MD5SUM_FILE = $(MD5SUM_DIR)/$(FILENAME).md5sum
+SCRIPT_DIR = $( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
+include $(shell rospack find mk)/download_unpack_build.mk
+INSTALL_DIR =	'./'
+
+
+installed.voicevox_model: $(SOURCE_DIR)/unpacked
+	mkdir -p $(INSTALL_DIR)/lib
+	cp build/$(DIRNAME)/lib/* $(INSTALL_DIR)/lib
+
+clean:
+	rm -rf $(TARBALL)
+	rm -rf $(SOURCE_DIR)
+	rm -rf $(INSTALL_DIR)/lib
+	rm -rf build

3rdparty/voicevox/Makefile.open_jtalk_dic

@@ -0,0 +1,25 @@
+# -*- makefile -*-
+
+all: installed.open_jtalk_dic
+
+VERSION = 1.11.1
+FILENAME = open_jtalk_dic_utf_8-1.11.tar.gz
+TARBALL = build/$(FILENAME)
+TARBALL_URL = "https://github.com/r9y9/open_jtalk/releases/download/v$(VERSION)/$(FILENAME)"
+SOURCE_DIR = build/open_jtalk_dic_utf_8-1.11
+UNPACK_CMD = tar xvzf
+MD5SUM_DIR = $(CURDIR)/md5sum
+MD5SUM_FILE = $(MD5SUM_DIR)/open_jtalk_dic.tar.gz.md5sum
+include $(shell rospack find mk)/download_unpack_build.mk
+INSTALL_DIR =	'./'
+
+
+installed.open_jtalk_dic: $(SOURCE_DIR)/unpacked
+	mkdir -p $(INSTALL_DIR)/dict
+	cp -r build/open_jtalk_dic_utf_8-1.11 $(INSTALL_DIR)/dict
+
+clean:
+	rm -rf $(TARBALL)
+	rm -rf $(SOURCE_DIR)
+	rm -rf $(INSTALL_DIR)/dict/open_jtalk_dic_utf_8-1.11
+	rm -rf build

3rdparty/voicevox/README.md

@@ -0,0 +1,103 @@
+# voicevox
+
+ROS Interface for [VOICEVOX](https://voicevox.hiroshiba.jp/) (AI speech synthesis)
+
+## TERM
+
+[VOICEVOX](https://voicevox.hiroshiba.jp/) is basically free to use, but please check the terms of use below.
+
+[TERM](https://voicevox.hiroshiba.jp/term)
+
+Each voice synthesis character has its own rules. Please use this package according to those terms.
+
+| Character name  |  term link |
+| ---- | ---- |
+| 四国めたん | https://zunko.jp/con_ongen_kiyaku.html |
+| ずんだもん | https://zunko.jp/con_ongen_kiyaku.html |
+| 春日部つむぎ | https://tsukushinyoki10.wixsite.com/ktsumugiofficial/利用規約 |
+| 波音リツ | http://canon-voice.com/kiyaku.html |
+| 雨晴はう | https://amehau.com/?page_id=225 |
+| 玄野武宏 | https://virvoxproject.wixsite.com/official/voicevoxの利用規約 |
+| 白上虎太郎 | https://virvoxproject.wixsite.com/official/voicevoxの利用規約 |
+| 青山龍星 | https://virvoxproject.wixsite.com/official/voicevoxの利用規約 |
+| 冥鳴ひまり | https://kotoran8zunzun.wixsite.com/my-site/利用規約 |
+| 九州そら | https://zunko.jp/con_ongen_kiyaku.html |
+
+## Installation
+
+Build this package.
+
+```bash
+cd /path/to/catkin_workspace
+catkin build voicevox
+```
+
+## Usage
+
+### Launch sound_play with VOICEVOX Text-to-Speech
+
+```bash
+roslaunch voicevox voicevox_texttospeech.launch
+```
+
+<a id="saysomething"></a>
+### Say something
+
+#### For python users
+
+```python
+import rospy
+from sound_play.libsoundplay import SoundClient
+
+rospy.init_node('say_node')
+
+client = SoundClient(sound_action='robotsound_jp', sound_topic='robotsound_jp')
+
+client.say('こんにちは', voice='四国めたん-あまあま')
+```
+
+You can change the voice by changing the voice_name.
+You can also specify the speaker id.
+Look at the following tables for further details.
+
+|  speaker_id  |  voice_name  |
+| ---- | ---- |
+| 0 | 四国めたん-あまあま |
+| 1 | ずんだもん-あまあま |
+| 2 | 四国めたん-ノーマル |
+| 3 | ずんだもん-ノーマル |
+| 4 | 四国めたん-セクシー |
+| 5 | ずんだもん-セクシー |
+| 6 | 四国めたん-ツンツン |
+| 7 | ずんだもん-ツンツン |
+| 8 | 春日部つむぎ-ノーマル |
+| 9 | 波音リツ-ノーマル |
+| 10 | 雨晴はう-ノーマル |
+| 11 | 玄野武宏-ノーマル |
+| 12 | 白上虎太郎-ノーマル |
+| 13 | 青山龍星-ノーマル |
+| 14 | 冥鳴ひまり-ノーマル |
+| 15 | 九州そら-あまあま |
+| 16 | 九州そら-ノーマル |
+| 17 | 九州そら-セクシー |
+| 18 | 九州そら-ツンツン |
+| 19 | 九州そら-ささやき |
+
+#### For roseus users
+
+```
+$ roseus
+(load "package://pr2eus/speak.l")
+
+(ros::roseus "say_node")
+
+(speak "JSKへようこそ。" :lang "波音リツ" :wait t :topic-name "robotsound_jp")
+```
+
+### Tips
+
+Normally, the server for speech synthesis starts up at `http://localhost:50021`.
+You can change the url and port by setting values for `VOICEVOX_TEXTTOSPEECH_URL` and `VOICEVOX_TEXTTOSPEECH_PORT`.
+
+You can also set the default character by setting `VOICEVOX_DEFAULT_SPEAKER_ID`.
+Please refer to [here](#saysomething) for the speaker id.