Speech & Image Recognizer for M5StickV(要ファームウェアの最小構成ビルド)

Speech_and_face_recognizer.py

import sensor
import image
import lcd
import KPU as kpu
import json
import time

from fpioa_manager import *
from machine import UART
from Maix import I2S, GPIO
import struct
from modules import SpeechRecognizer
import _thread

fm.register(35, fm.fpioa.UART2_TX, force=True)
fm.register(34, fm.fpioa.UART2_RX, force=True)

uart_Port = UART(UART.UART2, 1500000, 8, None, 1, timeout=1000, read_buf_len= 4096)

# -------------------- Button and Led
fm.register(board_info.BUTTON_A, fm.fpioa.GPIO1)
button_a = GPIO(GPIO.GPIO1, GPIO.IN, GPIO.PULL_UP)

fm.register(board_info.BUTTON_B, fm.fpioa.GPIO2)
button_b = GPIO(GPIO.GPIO2, GPIO.IN, GPIO.PULL_UP)

fm.register(board_info.LED_W, fm.fpioa.GPIO3)
led_w = GPIO(GPIO.GPIO3, GPIO.OUT)
led_w.value(1) #RGBW LEDs are Active Low

fm.register(board_info.LED_R, fm.fpioa.GPIO4)
led_r = GPIO(GPIO.GPIO4, GPIO.OUT)
led_r.value(1) #RGBW LEDs are Active Low

fm.register(board_info.LED_G, fm.fpioa.GPIO5)
led_g = GPIO(GPIO.GPIO5, GPIO.OUT)
led_g.value(1) #RGBW LEDs are Active Low

fm.register(board_info.LED_B, fm.fpioa.GPIO6)
led_b = GPIO(GPIO.GPIO6, GPIO.OUT)
led_b.value(1) #RGBW LEDs are Active Low

lcd.init(type=3)
lcd.rotation(2)
img = image.Image()
img.draw_string(0, 0, "FaceDetect\nDemo", scale=2, color=(255,255,255), x_spacing=1, mono_space=1)
led_w.value(0)
lcd.display(img)
time.sleep(3)
led_w.value(1)

# --------------------- speech Initialize

def sr_data_load(s, keyword_num, model_num, frame_num, path):
    # print(path)
    with open(path, 'r') as f:
        data = f.read()
        s.add_voice_model(keyword_num, model_num, frame_num, data)
def sr_data_allload():
    sr_data_load(s, 0, 0, 33, "/flash/sr/0_0.sr")
    sr_data_load(s, 0, 1, 19, "/flash/sr/0_1.sr")
    sr_data_load(s, 0, 2, 69, "/flash/sr/0_2.sr")
    sr_data_load(s, 1, 0, 22, "/flash/sr/1_0.sr")
    sr_data_load(s, 1, 1, 20, "/flash/sr/1_1.sr")
    sr_data_load(s, 1, 2, 20, "/flash/sr/1_2.sr")
    sr_data_load(s, 2, 0, 27, "/flash/sr/2_0.sr")
    sr_data_load(s, 2, 1, 22, "/flash/sr/2_1.sr")
    sr_data_load(s, 2, 2, 29, "/flash/sr/2_2.sr")
    print("load successful!")

print("voice recognize Init")
# register i2s(i2s0) pin
fm.register(board_info.MIC_DAT, fm.fpioa.I2S0_IN_D0, force=True)
fm.register(board_info.MIC_CLK, fm.fpioa.I2S0_SCLK, force=True)
fm.register(board_info.MIC_LRCLK, fm.fpioa.I2S0_WS, force=True)

sample_rate = 16000
# init i2s(i2s0)
i2s_dev = I2S(I2S.DEVICE_0)

# config i2s according to speechrecognizer
i2s_dev.channel_config(i2s_dev.CHANNEL_0,
                       I2S.RECEIVER,
                       resolution=I2S.RESOLUTION_16_BIT,
                       cycles=I2S.SCLK_CYCLES_32,
                       align_mode=I2S.STANDARD_MODE)
i2s_dev.set_sample_rate(sample_rate)
print("------")
s = SpeechRecognizer(i2s_dev)
print("------")
s.set_threshold(0, 0, 15000)  # 设置所处环境的噪声阈值, 环境噪声越大设置最后一个参数越

# --------------------- speech Initialize end


def image_recognize():
    print("image recognize Init")
    sensor.reset()
    sensor.set_pixformat(sensor.RGB565)
    sensor.set_framesize(sensor.QVGA)
    sensor.set_vflip(False)
    sensor.run(1)
    task = kpu.load(0x300000) # you need put model(face.kfpkg) in flash at address 0x300000
    anchor = (0.57273, 0.677385, 1.87446, 2.06253, 3.33843, 5.47434, 7.88282, 3.52778, 9.77052, 9.16828)
    a = kpu.init_yolo2(task, 0.5, 0.3, 5, anchor)
    print("image recognize End")
    while True:
        img = sensor.snapshot()
        img = img.resize(224,224)
        a = img.pix_to_ai()
        code = kpu.run_yolo2(task, img)
        if code:
            for i in code:
                print(i)
                print(type(i))
                a = img.draw_rectangle(i.rect())#, fill=True)
                uart_Port.write(json.dumps(i))
                print("Send")
        time.sleep_ms(100)
        a = lcd.display(img)


def voice_recognize():
    sr_data_allload()
    while True:
        # recognize
        s.recognize()
        ret = s.get_result()
        if ret > 0:
            if ret == 1:
                print("ret:{}-{}".format(ret, "homura"))
                led_r.value(0)
                time.sleep(1)
                led_r.value(1)
            elif ret == 2:
                print("ret:{}-{}".format(ret, "konbanwa"))
                led_g.value(0)
                time.sleep(1)
                led_g.value(1)
            elif ret == 3:
                print("ret:{}-{}".format(ret, "ohayou"))
                led_b.value(0)
                time.sleep(1)
                led_b.value(1)
        time.sleep_ms(100)

_thread.start_new_thread(image_recognize,())
_thread.start_new_thread(voice_recognize,())

while True:
    time.sleep(1)