Winston-503

from tensorflow.keras.applications import ResNet50

img_height, img_width = 250, 250  # size of images

base_model = ResNet50(weights='imagenet', 
                      include_top = False, 
                      # without dense part of the network
                      input_shape = (img_height, img_width, 3))

Winston-503

import speech_recognition as sr 

audio_file = sr.AudioFile('russian_audio.wav')
r = sr.Recognizer()

with audio_file as af:
    audio = r.record(af)

text = r.recognize_sphinx(audio, language='ru-RU')
print(f"Sphinx thinks you said:\n {text}")

Winston-503

import speech_recognition as sr 

italian_audio = sr.AudioFile('italian_audio.wav')
r = sr.Recognizer()

with italian_audio as af:
    audio = r.record(af)

text = r.recognize_google(audio, language='it-IT')
print(f"Google thinks you said:\n {text}")