app.py

import os
import io
from PIL import Image
from transformers import AutoImageProcessor, AutoModelForObjectDetection
import streamlit as st
import torch
import requests

#def prettier(results):
#    for item in results:
#        score = round(item['score'], 3)
 #       label = item['label']  # Use square brackets to access the 'label' key
  #      location = [round(value, 2) for value in item['box'].values()]
   #     print(f'Detected {label} with confidence {score} at location {location}')

def prettify_results(results):
    for item in results:
        score = round(item['score'].item(), 3)
        label = model.config.id2label[item['label']]  # Get label from id2label mapping in model config
        box = [round(coord, 2) for coord in item['box']]
        st.write(f'Detected {label} with confidence {score} at location {box}')
# Function to process uploaded image and prepare input for model

def input_image_setup(uploaded_file):
    if uploaded_file is not None:
        bytes_data = uploaded_file.getvalue()
        image = Image.open(io.BytesIO(bytes_data))  # Convert bytes data to PIL image
        return image
    else:
        raise FileNotFoundError("No file uploaded")


#Streamlit App
st.set_page_config(page_title="Image Detection")
st.header("Object Detection Application")
#Select your model
models = ["facebook/detr-resnet-50", "ciasimbaya/ObjectDetection", "hustvl/yolos-tiny"]  # List of supported models
model_name = st.selectbox("Select model", models)
processor = AutoImageProcessor.from_pretrained(model_name)
model = AutoModelForObjectDetection.from_pretrained(model_name)
#Upload an image
uploaded_file = st.file_uploader("choose an image...", type=["jpg","jpeg","png"])
image=""
if uploaded_file is not None:
    image = Image.open(uploaded_file)
    st.image(image, caption="Uploaded Image.", use_column_width=True)
submit = st.button("Detect Objects ")
if submit:
    image_data=input_image_setup(uploaded_file)
    st.subheader("The response is..")
    #process with model
    inputs = processor(images=image, return_tensors="pt")
    outputs = model(**inputs)

    # model predicts bounding boxes and corresponding COCO classes
    logits = outputs.logits
    bboxes = outputs.pred_boxes
    # print results
    target_sizes = torch.tensor([image.size[::-1]])
    results = processor.post_process_object_detection(outputs, threshold=0.9, target_sizes=target_sizes)[0]
    prettify_results(results)
#for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
 #       box = [round(i, 2) for i in box.tolist()]
  #      print(
   #         f"Detected {model.config.id2label[label.item()]} with confidence "
    #        f"{round(score.item(), 3)} at location {box}"
     #   )