Spaces:
Sleeping
Sleeping
Update src/streamlit_app.py
Browse files- src/streamlit_app.py +20 -20
src/streamlit_app.py
CHANGED
|
@@ -2,40 +2,40 @@ import streamlit as st
|
|
| 2 |
from transformers import pipeline
|
| 3 |
from PIL import Image
|
| 4 |
|
| 5 |
-
# Creates a brief description
|
| 6 |
def generate_caption(image):
|
| 7 |
-
with st.spinner("
|
| 8 |
# Loads the BLIP model to examine and describe the picture
|
| 9 |
-
image_to_text = pipeline("image-to-text", model="Salesforce/blip-image-captioning-
|
| 10 |
caption = image_to_text(image)[0]["generated_text"]
|
| 11 |
return caption
|
| 12 |
|
| 13 |
# Builds a story from the picture’s description
|
| 14 |
def generate_story(caption):
|
| 15 |
-
with st.spinner("
|
| 16 |
-
# Uses
|
| 17 |
-
pipe = pipeline("text-generation", model="
|
| 18 |
story = pipe(caption)[0]['generated_text']
|
| 19 |
return story
|
| 20 |
|
| 21 |
-
# Turns the story into
|
| 22 |
def generate_audio(story):
|
| 23 |
-
with st.spinner("
|
| 24 |
-
# Uses a speech model to
|
| 25 |
-
pipe = pipeline("text-to-speech", model="
|
| 26 |
audio = pipe(story)
|
| 27 |
return audio
|
| 28 |
|
| 29 |
-
# Streamlit UI: Makes a simple interface
|
| 30 |
|
| 31 |
-
# Displays
|
| 32 |
-
st.title("
|
| 33 |
|
| 34 |
-
# Describes the app for
|
| 35 |
-
st.write("
|
| 36 |
|
| 37 |
# Allows picture uploads
|
| 38 |
-
uploaded_file = st.file_uploader("
|
| 39 |
|
| 40 |
if uploaded_file is not None:
|
| 41 |
# Shows the uploaded picture
|
|
@@ -44,17 +44,17 @@ if uploaded_file is not None:
|
|
| 44 |
|
| 45 |
# Gets the picture’s description
|
| 46 |
image_caption = generate_caption(image)
|
| 47 |
-
st.subheader("
|
| 48 |
st.write(image_caption)
|
| 49 |
|
| 50 |
-
#
|
| 51 |
story_telling = generate_story(image_caption)
|
| 52 |
-
st.subheader("
|
| 53 |
st.write(story_telling)
|
| 54 |
|
| 55 |
# Generates audio
|
| 56 |
audio = generate_audio(story_telling)
|
| 57 |
-
if st.button("Hear
|
| 58 |
st.audio(audio['audio'],
|
| 59 |
format="audio/wav",
|
| 60 |
start_time=0,
|
|
|
|
| 2 |
from transformers import pipeline
|
| 3 |
from PIL import Image
|
| 4 |
|
| 5 |
+
# Creates a brief description for the pictures
|
| 6 |
def generate_caption(image):
|
| 7 |
+
with st.spinner("Analysing the Pictures for Key Message..."):
|
| 8 |
# Loads the BLIP model to examine and describe the picture
|
| 9 |
+
image_to_text = pipeline("image-to-text", model="Salesforce/blip-image-captioning-large")
|
| 10 |
caption = image_to_text(image)[0]["generated_text"]
|
| 11 |
return caption
|
| 12 |
|
| 13 |
# Builds a story from the picture’s description
|
| 14 |
def generate_story(caption):
|
| 15 |
+
with st.spinner("Enhancing the Story for better Details..."):
|
| 16 |
+
# Uses the text generation model to create a story based on the description
|
| 17 |
+
pipe = pipeline("text-generation", model="TheBloke/phi-2-GGUF")
|
| 18 |
story = pipe(caption)[0]['generated_text']
|
| 19 |
return story
|
| 20 |
|
| 21 |
+
# Turns the story into audio
|
| 22 |
def generate_audio(story):
|
| 23 |
+
with st.spinner("Turning story into News audio..."):
|
| 24 |
+
# Uses a speech model to turn description into audio
|
| 25 |
+
pipe = pipeline("text-to-speech", model="hexgrad/Kokoro-82M")
|
| 26 |
audio = pipe(story)
|
| 27 |
return audio
|
| 28 |
|
| 29 |
+
# Streamlit UI: Makes a simple interface to generate the audio
|
| 30 |
|
| 31 |
+
# Displays the title
|
| 32 |
+
st.title("Tool for the Reporter - Turning the News Photo into Audio")
|
| 33 |
|
| 34 |
+
# Describes the app for users
|
| 35 |
+
st.write("Please upload the News Photo within 200MB")
|
| 36 |
|
| 37 |
# Allows picture uploads
|
| 38 |
+
uploaded_file = st.file_uploader("Upload the Photo below", type=["png", "jpg", "jpeg"])
|
| 39 |
|
| 40 |
if uploaded_file is not None:
|
| 41 |
# Shows the uploaded picture
|
|
|
|
| 44 |
|
| 45 |
# Gets the picture’s description
|
| 46 |
image_caption = generate_caption(image)
|
| 47 |
+
st.subheader("Phot Description:")
|
| 48 |
st.write(image_caption)
|
| 49 |
|
| 50 |
+
# Generate the News descriptions
|
| 51 |
story_telling = generate_story(image_caption)
|
| 52 |
+
st.subheader("The News:")
|
| 53 |
st.write(story_telling)
|
| 54 |
|
| 55 |
# Generates audio
|
| 56 |
audio = generate_audio(story_telling)
|
| 57 |
+
if st.button("Hear the News"):
|
| 58 |
st.audio(audio['audio'],
|
| 59 |
format="audio/wav",
|
| 60 |
start_time=0,
|