source-separation-models

audio_example.wav is converted from
https://github.com/deezer/spleeter/blob/v1.4.0/audio_example.mp3
with the following command

sox audio_example.mp3 audio_example.wav

v1.12.0

What's Changed

• Fix building wheels for macOS by @csukuangfj in #2192
• Show verbose logs in homophone replacer by @csukuangfj in #2194
• Fix displaying streaming speech recognition results for Python. by @csukuangfj in #2196
• Add real-time speech recognition example for SenseVoice. by @csukuangfj in #2197
• docs: add Open-XiaoAI KWS project by @idootop in #2198
• Add C++ example for streaming ASR with SenseVoice. by @csukuangfj in #2199
• Add C++ example for real-time ASR with nvidia/parakeet-tdt-0.6b-v2. by @csukuangfj in #2201
• Add a link to YouTube video including sherpa-onnx. by @csukuangfj in #2202
• Support sending is_eof for online websocket server. by @csukuangfj in #2204
• Add alsa-based streaming ASR example for sense voice. by @csukuangfj in #2207
• Support homophone replacer in Android asr demo. by @csukuangfj in #2210
• Add a Go implementation of the TTS generation callback by @xiaokuang95 in #2213
• Add Android demo for real-time ASR with non-streaming ASR models. by @csukuangfj in #2214
• Expose dither for JNI by @esavin in #2215
• Add nodejs example for parakeet-tdt-0.6b-v2. by @csukuangfj in #2219
• Add script to build APK for simulated-streaming-asr. by @csukuangfj in #2220
• Release v1.12.0 by @csukuangfj in #2221

New Contributors

• @idootop made their first contribution in #2198
• @esavin made their first contribution in #2215

Full Changelog: v1.11.5...v1.12.0

v1.11.5

What's Changed

• export parakeet-tdt-0.6b-v2 to sherpa-onnx by @csukuangfj in #2180
• Add C++ runtime for parakeet-tdt-0.6b-v2. by @csukuangfj in #2181
• Avoid NaN in feature normalization. by @csukuangfj in #2186
• Release v1.11.5 by @csukuangfj in #2187

Full Changelog: v1.11.4...v1.11.5

v1.11.4

What's Changed

• Disable strict hotword matching mode for offline transducer by @vsd-vector in #1837
• Comment refinement: Add note about vocoder file for matcha TTS config by @HaoWang0101 in #2106
• Fix a typo in the JNI for Android. by @csukuangfj in #2108
• Generate subtitles with FireRedAsr models by @csukuangfj in #2112
• Use manylinux_2_28_x86_64 to build linux gpu for sherpa-onnx by @csukuangfj in #2123
• Support running sherpa-onnx with RK NPU on Android by @csukuangfj in #2124
• Fix building for HarmonyOS by @csukuangfj in #2125
• cmake build, configurable from env by @KarelVesely84 in #2115
• Expose dither in python API by @nshmyrev in #2127
• Add support for GigaAM-CTC-v2 by @rominf in #2135
• Support Giga AM transducer V2 by @csukuangfj in #2136
• Export kokoro 1.0 int8 models by @csukuangfj in #2137
• Upload more onnx ASR models by @csukuangfj in #2141
• Fix building for open harmonyOS by @csukuangfj in #2142
• online-transducer: reset the encoder toghter with 2 previous output symbols (non-blank) by @KarelVesely84 in #2129
• Fix punctuations for kokoro tts 1.1-zh. by @csukuangfj in #2146
• Fix setting OnlineModelConfig in Java API by @csukuangfj in #2147
• Support decoding multiple streams in Java API. by @csukuangfj in #2149
• Support replacing homonphonic phrases by @csukuangfj in #2153
• Add C and CXX API for homophone replacer by @csukuangfj in #2156
• Add JavaScript API (WASM) for homophone replacer by @csukuangfj in #2157
• Add JavaScript API (node-addon) for homophone replacer by @csukuangfj in #2158
• Fix building without TTS by @csukuangfj in #2159
• Add homonphone replacer example for Python API. by @csukuangfj in #2161
• More fix for building without tts by @csukuangfj in #2162
• Add Swift API for homophone replacer. by @csukuangfj in #2164
• Add C# API for homophone replacer by @csukuangfj in #2165
• Add Kotlin and Java API for homophone replacer by @csukuangfj in #2166
• Add Dart API for homophone replacer by @csukuangfj in #2167
• Add Go API for homophone replacer by @csukuangfj in #2168
• Release v1.11.4 by @csukuangfj in #2169

New Contributors

• @HaoWang0101 made their first contribution in #2106

Full Changelog: v1.11.3...v1.11.4

hr-files

replace.fst is generated from
https://colab.research.google.com/drive/1jEaS3s8FbRJIcVQJv2EQx19EM_mnuARi?usp=sharing

---

If you don't have access to the colab notebook, here is the code for generating replace.fst:

import pynini
from pynini.lib import utf8, byte
from pynini import cdrewrite

sigma = utf8.VALID_UTF8_CHAR.star

rule1 = pynini.cross("dan1ni2er3bo1wei2", "丹尼尔·波维")
rule10 = pynini.cross("dan1ni2er3bo1wei4", "丹尼尔·波维")
rule2 = pynini.cross('dou4dou4', '豆豆')
rule3 = pynini.cross('cheng2cheng2', '橙橙')
rule30 = pynini.cross('chen2chen2', '橙橙')
rule4 = pynini.cross('qiao2qiao2', '峤峤')
rule5 = pynini.cross('qiu2qiu2', '球球')
rule6 = pynini.cross('lin2mei3li4', '林美丽')
rule7 = pynini.cross('guo3guo3', '果果')
rule8 = pynini.cross('miao2miao2', '苗苗')
rule9 = pynini.cross('xuan2jie4', '玄戒')
rule10 = pynini.cross('xuan2jie4xin1pian1', '玄戒芯片')
rule11 = pynini.cross('xuan2jie4xing1pian1', '玄戒芯片')
rule12 = pynini.cross('xuan2jie4xin1pian1', '玄戒芯片')
rule13 = pynini.cross('xuan2jie4xing1pian1', '玄戒芯片')


rule = (rule1 | rule10 | rule2 | rule3 | rule30 | rule4 | rule5 | rule6 | rule7 | rule8 | rule9 | rule10 | rule11 | rule12 | rule13).optimize()
rule = cdrewrite(rule, "", "", sigma)

rule.write('replace.fst')

Note that you need to use

pip install --only-binary :all: pynini

to install pynini

v1.11.3

What's Changed

• fix vits dict dir config by @amutu in #2036
• fix case by @amutu in #2037
• Fix building wheels for RKNN by @csukuangfj in #2041
• 缩放因子应该是32767？ by @yourengod in #2056
• Fix length scale for kokoro tts by @csukuangfj in #2060
• Allow building repository as CMake subdirectory by @niansa in #2059
• Export silero_vad v4 to RKNN by @csukuangfj in #2067
• 修复 DirectML 支持 by @endink in #2066
• Fix building aar to include speech denoiser by @csukuangfj in #2069
• Add CXX API for VAD by @csukuangfj in #2077
• Add C++ runtime for silero_vad with RKNN by @csukuangfj in #2078
• Refactor rknn code by @csukuangfj in #2079
• Fix building for android by @csukuangfj in #2081
• Add C++ and Python API for Dolphin CTC models by @csukuangfj in #2085
• Add Kotlin and Java API for Dolphin CTC models by @csukuangfj in #2086
• Add C and CXX API for Dolphin CTC models by @csukuangfj in #2088
• Preserve more context after endpointing in transducer by @vsd-vector in #2061
• Add C# API for Dolphin CTC models by @csukuangfj in #2089
• Add Go API for Dolphin CTC models by @csukuangfj in #2090
• Add Swift API for Dolphin CTC models by @csukuangfj in #2091
• Add Javascript (WebAssembly) API for Dolphin CTC models by @csukuangfj in #2093
• Add Javascript (node-addon) API for Dolphin CTC models by @csukuangfj in #2094
• Add Dart API for Dolphin CTC models by @csukuangfj in #2095
• Add Pascal API for Dolphin CTC models by @csukuangfj in #2096
• Release v1.11.3 by @csukuangfj in #2097

New Contributors

• @amutu made their first contribution in #2036
• @yourengod made their first contribution in #2056
• @niansa made their first contribution in #2059

Full Changelog: v1.11.2...v1.11.3

v1.11.2

What's Changed

• Fix CI tests. by @csukuangfj in #2016
• Publish jar for more java versions by @csukuangfj in #2017
• add alsa example for vad+offline asr by @csukuangfj in #2020
• Support cuda12 and cudnn8 for Linux aarch64. by @csukuangfj in #2021
• Update README to include more projects using sherpa-onnx by @csukuangfj in #2022
• Fix a bug in vad.reset() by @csukuangfj in #2023
• Fix Matcha + vocos for Android by @csukuangfj in #2024
• Fix crash in Android tts engine demo. by @csukuangfj in #2029
• Fix build script by @sienaiwun in #2033
• fix static linking by @sangeet2020 in #2032
• Release v1.11.2 by @csukuangfj in #2035

New Contributors

• @sienaiwun made their first contribution in #2033

Full Changelog: v1.11.1...v1.11.2

v1.11.1

What's Changed

• Export vocos to sherpa-onnx by @csukuangfj in #2012
• Add C++ runtime for vocos by @csukuangfj in #2014
• Release v1.11.1 by @csukuangfj in #2015

Full Changelog: v1.11.0...v1.11.1

v1.11.0

What's Changed

• Fix building wheels for Python 3.7 by @csukuangfj in #1933
• Add Kotlin and Java API for online punctuation models by @csukuangfj in #1936
• Add Kokoro v1.1-zh by @csukuangfj in #1942
• Support RKNN for Zipformer CTC models. by @csukuangfj in #1948
• Add transducer modified_beam_search for RKNN. by @csukuangfj in #1949
• Update README to include projects that is using sherpa-onnx by @csukuangfj in #1956
• Limit number of tokens per second for whisper. by @csukuangfj in #1958
• Ebranchformer by @KarelVesely84 in #1951
• Test using sherpa-onnx as a cmake subproject by @csukuangfj in #1961
• Add C++ demo for VAD+non-streaming ASR by @csukuangfj in #1964
• Export gtcrn models to sherpa-onnx by @csukuangfj in #1975
• c-api add wave write to buffer. by @cjsdurj in #1962
• add SherpaOnnxOfflineRecognizerSetConfig binding for go by @franck-li in #1976
• Add C++ runtime for speech enhancement GTCRN models by @csukuangfj in #1977
• Add Python API for speech enhancement GTCRN models by @csukuangfj in #1978
• Add C API for speech enhancement GTCRN models by @csukuangfj in #1984
• Add CXX API for speech enhancement GTCRN models by @csukuangfj in #1986
• Add Swift API for speech enhancement GTCRN models by @csukuangfj in #1989
• Add C# API for speech enhancement GTCRN models by @csukuangfj in #1990
• Add Go API for speech enhancement GTCRN models by @csukuangfj in #1991
• Add Pascal API for speech enhancement GTCRN models by @csukuangfj in #1992
• Add Dart API for speech enhancement GTCRN models by @csukuangfj in #1993
• Add JavaScript (node-addon) API for speech enhancement GTCRN models by @csukuangfj in #1996
• Add WebAssembly (WASM) for speech enhancement GTCRN models by @csukuangfj in #2002
• Add JavaScript API (wasm) for speech enhancement GTCRN models by @csukuangfj in #2007
• Add Kotlin API for speech enhancement GTCRN models by @csukuangfj in #2008
• Add Java API for speech enhancement GTCRN models by @csukuangfj in #2009
• Release v1.11.0 by @csukuangfj in #2010

New Contributors

• @cjsdurj made their first contribution in #1962

Full Changelog: v1.10.46...v1.11.0

speech-enhancement-models

gtrcn_simple.onnx is from https://github.com/Xiaobin-Rong/gtcrn

speech_with_noise.wav is from https://modelscope.cn/models/iic/speech_zipenhancer_ans_multiloss_16k_base/file/view/master?fileName=examples%252Fspeech_with_noise.wav&status=0