main.py

from transformers import DetrFeatureExtractor, DetrForObjectDetection
import torch
from PIL import Image
from fastapi import FastAPI,Response,UploadFile,File
from io import BytesIO


app = FastAPI()

@app.post("/")
async def postImg(file: bytes = File()):
    if not file:
        return {"message": "No file sent"}
    else:
        objects = []
        image = Image.open(BytesIO(file))
        feature_extractor = DetrFeatureExtractor.from_pretrained("facebook/detr-resnet-50")
        model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50")

        inputs = feature_extractor(images=image, return_tensors="pt")
        outputs = model(**inputs)

        # convert outputs (bounding boxes and class logits) to COCO API
        target_sizes = torch.tensor([image.size[::-1]])
        results = feature_extractor.post_process(outputs, target_sizes=target_sizes)[0]

        for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
            box = [round(i, 2) for i in box.tolist()]
            # let's only keep detections with score > 0.9
            if score > 0.9:
                objects.append(model.config.id2label[label.item()])
        return {"objects":objects}