app.py

import torch
import streamlit as st
import numpy as np
from PIL import Image
from transformers import pipeline
from transformers import OwlViTProcessor, OwlViTForObjectDetection
from tempfile import NamedTemporaryFile

audiopipe = pipeline("automatic-speech-recognition", model="openai/whisper-large-v3")
#imagepipe = pipeline("image-classification", model="Kaludi/food-category-classification-v2.0")
#imagepipe = pipeline("image-classification", model="nateraw/food")
imagepipe = pipeline("image-classification", model="flatmoon102/fruits_and_vegetables_image_classification")

processor = OwlViTProcessor.from_pretrained("google/owlvit-base-patch32")
model = OwlViTForObjectDetection.from_pretrained("google/owlvit-base-patch32")

st.title('Upload an audio file for speech recognition')

uploaded_audio_file = st.file_uploader("Choose an audio file (wav)")
if uploaded_audio_file is not None:
    with NamedTemporaryFile(suffix="wav") as temp:
        temp.write(uploaded_audio_file.getvalue())
        temp.seek(0)
        result = audiopipe(temp.name)
        st.write(result)

st.title('Upload an image file to classification (food)')

uploaded_image_file = st.file_uploader("Choose an image file")
if uploaded_image_file is not None:
    with NamedTemporaryFile() as temp:
        temp.write(uploaded_image_file.getvalue())
        temp.seek(0)
        result = imagepipe(temp.name)
        st.write(result)

st.title('Upload an image file to detection')

uploaded_image_zero_file = st.file_uploader("Choose an image file (zero)")
texts = st.text_input('tags')
if uploaded_image_zero_file is not None:
    image = Image.open(uploaded_image_zero_file)
    outputImage = np.array(image)
    st.image(outputImage)

if st.button('apply tag'):
    tags = [texts.split(", ")]
    inputs = processor(text=tags, images=image, return_tensors="pt")
    outputs = model(**inputs)
    target_sizes = torch.Tensor([image.size[::-1]])
    results = processor.post_process_object_detection(outputs=outputs, threshold=0.1, target_sizes=target_sizes)
    i = 0  # Retrieve predictions for the first image for the corresponding text queries
    text = texts[i]
    boxes, scores, labels = results[i]["boxes"], results[i]["scores"], results[i]["labels"]
    st.write(results)
    # Print detected objects and rescaled box coordinates
    for box, score, label in zip(boxes, scores, labels):
        box = [round(i, 2) for i in box.tolist()]
        print(f"Detected {text[label]} with confidence {round(score.item(), 3)} at location {box}")