Add application file

app.py

@@ -0,0 +1,74 @@
+import streamlit as st
+from transformers import pipeline
+from gtts import gTTS 
+from PIL import Image
+import time
+#from playsound import playsound
+#from pydub import AudioSegment
+#from preferredsoundplayer import soundplay
+
+#@st.cache()
+@st.cache(allow_output_mutation=True)
+def load_model():
+    """Retrieves the trained model"""
+    model = pipeline('image-to-text')
+    return model
+
+def main():
+	caption = load_model()
+	st.title("Welcome to image to speech app")
+	instructions = """Click an image using inbuilt camera 
+	or upload an image file"""
+	st.write(instructions)
+
+	img = None
+	pictureCam = st.camera_input("Take a picture")
+	pictureUpload = st.file_uploader('Upload An Image')
+
+	if pictureCam :
+		st.write('clicked image from webcam')
+		st.image(pictureCam)
+		img = Image.open(pictureCam)
+	elif pictureUpload : 
+		st.write('uploaded image from device')
+		st.image(pictureUpload)
+		img = Image.open(pictureUpload)
+
+	if img is not None : 
+		description = caption(img)
+		generated_text = description[0]['generated_text']
+		st.write(generated_text)
+		generated_audio = gTTS(generated_text)
+		generated_audio.save('demo.mp3')
+
+    	audio_file = open(‘demo.mp3’, ‘rb’)
+		audio_bytes = audio_file.read()
+		st.audio(audio_bytes, format=‘audio/ogg’,start_time=0)
+
+		#html_string = """
+        #<audio controls autoplay>
+        #  <source src="demo.mp3" type="audio/mp3">
+        #</audio>
+        #"""
+		#sound = st.empty()
+		#sound.markdown(html_string, unsafe_allow_html=True)  # will display a st.audio with the sound you specified in the "src" of the html_string and autoplay it
+		#time.sleep(2)  # wait for 2 seconds to finish the playing of the audio
+		#sound.empty()  # optionally delete the element afterwards
+
+		#sound = AudioSegment.from_mp3("demo.mp3")
+		#sound.export("demo.wav", format="wav")
+		#soundplay("demo.mp3")
+		#playsound('demo.mp3')
+		#audio_file = open('demo.wav', 'rb')
+		#audio_bytes = audio_file.read()
+		#st.audio(audio_bytes, format='audio/wav')
+
+
+if __name__ == '__main__' : 
+	main()
+
+	import streamlit as st
+import time
+
+
+