mirror of
https://github.com/espressif/esp-sr.git
synced 2025-09-15 15:28:44 +08:00
93 lines
3.2 KiB
Markdown
93 lines
3.2 KiB
Markdown
## ESP Chinese TTS [[中文]](./README.md)
|
||
|
||
Espressif Chinese TTS is a lightweight TTS system designed for embedded systems。
|
||
|
||
## Overview
|
||
|
||
The Chinese TTS is based on concatenative method. The flow diagram of system is as follows:
|
||
|
||
|
||
|
||
- **Parser** : a Chinese grapheme to phoneme module, input text (UTF-8) and output Chinese pinyin list.
|
||
- **Synthesizer** : a concatenative synthesizer, input pinyin list and output wave raw data. The default encoding of raw data is mono, 16 bit@16000 Hz.
|
||
|
||
#### Features
|
||
|
||
- [x] UTF-8 encoding text input
|
||
|
||
- [x] Streaming output
|
||
|
||
- [x] Polyphonic pronunciation
|
||
|
||
- [x] Adjustable speech rate
|
||
|
||
- [x] Digital broadcasting optimization
|
||
|
||
- [ ] Custom sound set
|
||
|
||
|
||
|
||
## Performance Test
|
||
|
||
#### Resource Occupancy
|
||
|
||
Flash image size： 2.2 MB
|
||
|
||
RAM runtime: 20 KB
|
||
|
||
CPU loading test（ESP32 @ 240 MHz）:
|
||
|
||
| speech rate | 0 | 1 | 2 | 3 | 4 | 5 |
|
||
| --------------------------- | :--: | :--: | :--: | :--: | :--: | :--: |
|
||
| times faster than real time | 4.5 | 3.2 | 2.9 | 2.5 | 2.2 | 1.8 |
|
||
|
||
**Note:** the bigger rate, the faster speech speed. 0: slowest speaking speed, 5: fastest speaking speed.
|
||
|
||
#### Samples
|
||
|
||
- 欢迎使用乐鑫语音合成, [voice=小乐,speed=0](./samples/S1_xiaole_speed0.wav), [voice=小乐,speed=2](./samples/S1_xiaole_speed2.wav)
|
||
|
||
- 支付宝收款 1111.11 元, [voice=小乐,speed=0](./samples/S1_xiaole_speed0.wav), [voice=小乐,speed=2](./samples/S2_xiaole_speed2.wav)
|
||
|
||
- 空调制热模式已打开，并调节到25度, [voice=小乐,speed=0](./samples/S3_xiaole_speed0.wav), [voice=小乐,speed=4](./samples/S3_xiaole_speed4.wav)
|
||
|
||
## User Guide
|
||
|
||
```c
|
||
#include "esp_tts.h"
|
||
#include "esp_tts_voice_female.h"
|
||
#include "esp_partition.h"
|
||
|
||
/*** 1. create esp tts handle ***/
|
||
|
||
//// Method1: use pre-define xiaole voice lib.
|
||
//// This method is not recommended because the method may make app bin exceed the limit of esp32
|
||
// esp_tts_handle_t *tts_handle=esp_tts_create(esp_tts_voice_female);
|
||
|
||
|
||
// method2: initial voice set from separate voice data partition
|
||
|
||
const esp_partition_t* part=esp_partition_find_first(ESP_PARTITION_TYPE_DATA, ESP_PARTITION_SUBTYPE_DATA_FAT, "voice_data");
|
||
if (part==0) printf("Couldn't find voice data partition!\n");
|
||
spi_flash_mmap_handle_t mmap;
|
||
uint16_t* voicedata;
|
||
esp_err_t err=esp_partition_mmap(part, 0, 3*1024*1024, SPI_FLASH_MMAP_DATA, (const void**)&voicedata, &mmap);
|
||
esp_tts_voice_t *voice=esp_tts_voice_set_init(&esp_tts_voice_template, voicedata);
|
||
|
||
// 2. parse text and synthesis wave data
|
||
char *text="欢迎使用乐鑫语音合成";
|
||
if (esp_tts_parse_chinese(tts_handle, text)) { // parse text into pinyin list
|
||
int len[1]={0};
|
||
do {
|
||
short *data=esp_tts_stream_play(tts_handle, len, 4); // streaming synthesis
|
||
i2s_audio_play(data, len[0]*2, portMAX_DELAY); // i2s output
|
||
} while(len[0]>0);
|
||
i2s_zero_dma_buffer(0);
|
||
}
|
||
|
||
```
|
||
|
||
please refer to [esp_tts.h](./esp_tts_chinese/include/esp_tts.h) for the details of API or [chinese_tts](../../examples/chinese_tts) example in esp-skainet.
|
||
|
||
|