speech_recognition/esp-sr
1
0
Fork 0
mirror of https://github.com/espressif/esp-sr.git synced 2025-09-15 15:28:44 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
ESP-SR 语音识别框架
259 Commits 23 Branches 16 Tags 323 MiB
C 98.1%
C++ 1.2%
Python 0.5%
CMake 0.2%
7f412e4c4b
Go to file
sxy 7f412e4c4b feat(wn9): Enable preload for wakenet9
2023-01-13 21:01:01 +08:00
.github/workflows ci: Add github upload component action 2022-12-14 13:57:49 +08:00
docs docs:fix_bugs 2022-12-02 15:13:10 +08:00
esp-tts bugfix(tts): Fix some wrong pronounce 2022-11-04 17:11:17 +08:00
include Update wakenet trigger v4 2023-01-07 15:33:58 +08:00
lib feat(wn9): Enable preload for wakenet9 2023-01-13 21:01:01 +08:00
model feat(wn9): Enable preload for wakenet9 2023-01-13 21:01:01 +08:00
src Update wakenet trigger v4 2023-01-07 15:33:58 +08:00
tool doc(tool): Rename the README 2021-11-18 21:03:40 +08:00
.gitignore testing 2022-11-23 20:01:24 +08:00
.gitlab-ci.yml ci: Push master and release branch to github 2022-12-14 11:09:12 +08:00
.readthedocs.yml doec/trans_md_2_rst 2022-11-15 17:18:30 +08:00
CHANGELOG.md feature/add AFE component & update wakenet 2021-02-25 11:01:29 +08:00
CMakeLists.txt bugfix: Remove Multiply definition 'print_afe_info' 2022-11-01 17:23:00 +08:00
component.mk bugfix: Add src/include path in component.mk 2022-08-17 20:57:55 +08:00
idf_component.yml ci: Add github upload component action 2022-12-14 13:57:49 +08:00
Kconfig.projbuild feat(wn9): Enable preload for wakenet9 2023-01-13 21:01:01 +08:00
libversion feat(MN): Fix the memory leak of MN; Modify the method of resetting speech commands 2022-02-17 14:23:09 +08:00
LICENSE add README 2019-08-06 12:01:02 +08:00
README.md docs: fix same bugs in docs 2022-12-02 11:39:21 +08:00

README.md

esp_sr

Espressif esp_sr provides basic algorithms for Speech Recognition applications. Now, this framework has four modules:

  • The wake word detection model WakeNet
  • The speech command recognition model MultiNet
  • Audio Front-End AFE
  • The text to speech model esp-tts

These algorithms are provided in the form of a component, so they can be integrated into your projects with minimum efforts. ESP32-S3 is recommended, which supports AI instructions and larger, high-speech octal SPI PSRAM. The new algorithms will no longer support ESP32 chips.

Wake Word Engine

Espressif wake word engine WakeNet is specially designed to provide a high performance and low memory footprint wake word detection algorithm for users, which enables devices always listen wake words, such as “Alexa”, “Hi,lexin” and “Hi,ESP”. You can refer to Model loading method to build your project.

Currently, Espressif has not only provided an official wake word "Hi,Lexin","Hi,ESP" to public for free, but also allows customized wake words. For details on how to customize your own wake words, please see Espressif Speech Wake Words Customization Process.

Speech Command Recognition

Espressif's speech command recognition model MultiNet is specially designed to provide a flexible off-line speech command recognition model. With this model, you can easily add your own speech commands, eliminating the need to train model again. You can refer to Model loading method to build your project.

Currently, Espressif MultiNet supports up to 200 Chinese or English speech commands, such as “打开空调” (Turn on the air conditioner) and “打开卧室灯” (Turn on the bedroom light).

Audio Front End

Espressif Audio Front-End AFE integrates AEC (Acoustic Echo Cancellation), VAD (Voice Activity Detection), BSS(Blind Source Separation) and NS (Noise Suppression).

Our two-mic Audio Front-End (AFE) have been qualified as a “Software Audio Front-End Solution” for Amazon Alexa Built-in devices.

In order to achieve optimal performance: