File size: 1,400 Bytes
c2fbc3b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from typing import Dict, List, Any
import os
import requests
from flask import Flask, Response, request, jsonify
from segment_anything import SamPredictor, sam_model_registry

class EndpointHandler():
    def __init__(self, path=""):
        # Preload all the elements you are going to need at inference.
        model_type = "vit_b"
		# prefix = "/opt/ml/model"
		model_path = "tf_model.h5"
		# model_checkpoint_path = os.path.join(prefix, "sam_vit_h_4b8939.pth")

		sam = sam_model_registry[model_type](checkpoint=model_path)
		predictor = SamPredictor(sam)

    def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
        """
       data args:
            inputs (:obj: `str` | `PIL.Image` | `np.array`)
            kwargs
      Return:
            A :obj:`list` | `dict`: will be serialized and returned
        """

		inputs = data.pop("inputs", data)
		image_url = inputs.pop("imageUrl", none)

		if not image_url:
			return jsonify({"error": "image_url not provided"}), 400

		try:
			response = requests.get(image_url)
			response.raise_for_status()
			image = response.content
		except requests.RequestException as e:
			return jsonify({"error": f"Error downloading image: {str(e)}"}), 500


		predictor.set_image(image)

		image_embedding = predictor.get_image_embedding().cpu().numpy().toList()

		return jsonify(image_embedding)

        # pseudo
        # self.model(input)