Google Teachable Machine 聲音分類結合 Linkit 7697，拍手彈指都聽得懂

python TM2_tflite_audio.py --model model.tflite --labels labels.txt

pip install tensorflow==2.7.0

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

# Imports
# from tflite_support.task import audio
# from tflite_support.task import core
# from tflite_support.task import processor

import argparse
import serial
import io
import time
import numpy as np
import cv2
import tensorflow as tf
import pyaudio
import wave

# from tflite_support.task import audio
# from tflite_support.task import core
# from tflite_support.task import processor
#import tensorflow as tf

from PIL import Image
# from tf.lite.interpreter import Interpreter

def load_labels(path):
  with open(path, 'r') as f:
    return {i: line.strip() for i, line in enumerate(f.readlines())}

def set_input_tensor(interpreter, image):
  tensor_index = interpreter.get_input_details()[0]['index']
  input_tensor = interpreter.tensor(tensor_index)()[0]
  input_tensor[:] = image

def main():
  BAUD_RATES = 9600
  # 修改實際的 com port
  ser = serial.Serial('com7', BAUD_RATES)
  while (True):
    parser = argparse.ArgumentParser(
        formatter_class=argparse.ArgumentDefaultsHelpFormatter)
    parser.add_argument(
        '--model', help='File path of .tflite file.', required=True)
    parser.add_argument(
        '--labels', help='File path of labels file.', required=True)
    args = parser.parse_args()
  
    labels = load_labels(args.labels)
  

    interpreter = tf.lite.Interpreter(args.model)
    # interpreter = Interpreter(args.model)
    
    interpreter.allocate_tensors()
    a = interpreter.get_input_details()[0]['shape']
    print(a)
    getaudio()
    x = np.fromfile(open('output.wav'),np.int32)[11:]

    # x[0, 0] = np.fromfile(open('output.wav'),np.int16)[22:]

    # for i in range(44032):
    #   x[0, i] = np.fromfile(open('output.wav'),np.int16)[22:][i]

    print(x.shape)
    results = classify_audio(interpreter, x)

    label_id, prob = results[0]
    print(labels[label_id],prob)

    if (labels[label_id]=='0 Background Noise') :
      ser.write(b'command_1\n')  # 訊息必須是位元組類型
      # time.sleep(0.5)           # 暫停0.5秒，再執行底下接收回應訊息的迴圈

    elif (labels[label_id]=='1 Clapping') :
      ser.write(b'command_2\n')  # 訊息必須是位元組類型
      # time.sleep(0.5)           # 暫停0.5秒，再執行底下接收回應訊息的迴圈

    elif (labels[label_id]=='2 Snapping') :
      ser.write(b'command_3\n')  # 訊息必須是位元組類型
      # time.sleep(0.5)           # 暫停0.5秒，再執行底下接收回應訊息的迴圈
    elif (labels[label_id]=='3 Sorry') :
      ser.write(b'command_4\n')  # 訊息必須是位元組類型
      # time.sleep(0.5)           # 暫停0.5秒，再執行底下接收回應訊息的迴圈
    time.sleep(2)
   
def classify_audio(interpreter, image, top_k=1):
  """Returns a sorted array of classification results."""
  set_input_tensor(interpreter, image)
  interpreter.invoke()
  output_details = interpreter.get_output_details()[0]
  output = np.squeeze(interpreter.get_tensor(output_details['index']))

  # If the model is quantized (uint8 data), then dequantize the results
  if output_details['dtype'] == np.uint8:
    scale, zero_point = output_details['quantization']
    output = scale * (output - zero_point)

  ordered = np.argpartition(-output, top_k)
  return [(i, output[i]) for i in ordered[:top_k]]

def getaudio():
  p = pyaudio.PyAudio()

  CHANNELS = 1
  FORMAT = pyaudio.paInt16
  CHUNK = 1024
  RATE = 44100
  p = pyaudio.PyAudio()

  stream = p.open(format=FORMAT,
                channels=CHANNELS,
                rate=RATE,
                input=True,
                frames_per_buffer=CHUNK)

  #print("Start recording!")

  frames = []
  seconds = 2
  for i in range (0, int(RATE/CHUNK*seconds)):
    data = stream.read(CHUNK)
    frames.append(data)

  #print("Stop recording.")
  stream.stop_stream()
  stream.close()

  p.terminate()

  wf = wave.open("output.wav", 'wb')
  wf.setnchannels(CHANNELS)
  wf.setsampwidth(p.get_sample_size(FORMAT))
  wf.setframerate(RATE)
  wf.writeframes(b''.join(frames))
  wf.close()

main()

#define LED_1_pin 2
#define LED_2_pin 3
#define LED_3_pin 4

String str;

void setup(void)  
{
  Serial.begin(9600);

  // init pin states
  pinMode(LED_1_pin, OUTPUT);
  digitalWrite(LED_1_pin,LOW);
  pinMode(LED_2_pin, OUTPUT);
  digitalWrite(LED_2_pin,LOW);
  pinMode(LED_3_pin, OUTPUT);
  digitalWrite(LED_3_pin,LOW);

}

void loop(void)  
{
  int i;
  
  if (Serial.available()) {
    // 讀取傳入的字串直到"\n"結尾
    str = Serial.readStringUntil('\n');

    if (str == "command_1") {     // 若字串值是 "command_1" 燈號OX
      digitalWrite(LED_1_pin,HIGH);
      digitalWrite(LED_2_pin,LOW);
      digitalWrite(LED_3_pin,LOW);

      Serial.println("command_1"); 
    } 
    else if (str == "command_2") {  // 若字串值是 "command_2" 燈號XO
      digitalWrite(LED_1_pin,LOW);
      digitalWrite(LED_2_pin,HIGH);
      digitalWrite(LED_3_pin,LOW);

      Serial.println("command_2");
    } 
    else if (str == "command_3" ) { // 若字串值是 "command_3" 燈號XX
      digitalWrite(LED_1_pin,LOW);  
      digitalWrite(LED_2_pin,LOW);
      digitalWrite(LED_3_pin,HIGH);

      Serial.println("command_3");
    }
    else if (str == "command_4" ) { // 若字串值是 "command_3" 燈號XX
      digitalWrite(LED_1_pin,HIGH);  
      digitalWrite(LED_2_pin,HIGH);
      digitalWrite(LED_3_pin,HIGH);

      Serial.println("command_4");
    }
  }
}