Spaces:
Runtime error
Runtime error
File size: 2,311 Bytes
45296db 149da67 45296db 5b2cffc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
#imported all required libraries
import streamlit as st
import torch
import requests
from PIL import Image
from io import BytesIO
from transformers import ViTFeatureExtractor, AutoTokenizer, VisionEncoderDecoderModel
#used a pretrained model hosted on huggingface
loc = "ydshieh/vit-gpt2-coco-en"
feature_extractor = ViTFeatureExtractor.from_pretrained(loc)
tokenizer = AutoTokenizer.from_pretrained(loc)
model = VisionEncoderDecoderModel.from_pretrained(loc)
model.eval()
#defined a function for prediction
def predict(image):
pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values
with torch.no_grad():
output_ids = model.generate(pixel_values, max_length=16, num_beams=4, return_dict_in_generate=True).sequences
preds = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
preds = [pred.strip() for pred in preds]
return preds
#defined a function for Streamlit App
def app():
st.title("ImaginateAI")
st.write("ViT and GPT2 are used to generate Image Caption for the uploaded image. COCO Dataset was used for training. This image captioning model might have some biases that I couldn’t figure during testing")
st.write("Upload an image or paste a URL to get predicted captions.")
upload_option = st.selectbox("Choose an option:", ("Upload Image", "Paste URL"))
if upload_option == "Upload Image":
uploaded_file = st.file_uploader("Upload an image", type=["jpg", "jpeg"])
if uploaded_file is not None:
image = Image.open(uploaded_file)
preds = predict(image)
st.image(image, caption="Uploaded Image", use_column_width=True)
st.write("Predicted Caption:", preds)
elif upload_option == "Paste URL":
image_url = st.text_input("Enter Image URL")
if st.button("Submit") and image_url:
try:
response = requests.get(image_url, stream=True)
image = Image.open(BytesIO(response.content))
preds = predict(image)
st.image(image, caption="Image from URL", use_column_width=True)
st.write("Predicted Caption:", preds)
except:
st.write("Error: Invalid URL or unable to fetch image.")
if __name__ == "__main__":
app() |