fastapi_es/image_vector.py at main · SimJunSik/fastapi_es · GitHub

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
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
import numpy as np
from keras.applications import InceptionResNetV2
from keras.applications.inception_resnet_v2 import preprocess_input
from keras.models import Model
from keras_preprocessing import image
from opensearchpy import OpenSearch, RequestsHttpConnection
from dotenv import load_dotenv
from requests_aws4auth import AWS4Auth

import os

load_dotenv(dotenv_path="./secrets/.env")


AWS_ACCESS_KEY = os.environ.get("AWS_ACCESS_KEY")
AWS_SECRET_KEY = os.environ.get("AWS_SECRET_KEY")
AWS_REGION = os.environ.get("AWS_REGION")
AWS_SERVICE = os.environ.get("AWS_SERVICE")

HOST = os.environ.get("HOST")

awsauth = AWS4Auth(
    AWS_ACCESS_KEY,
    AWS_SECRET_KEY,
    AWS_REGION,
    AWS_SERVICE,
)

es = OpenSearch(
    hosts=[{"host": HOST, "port": 443}],
    http_auth=awsauth,
    use_ssl=True,
    verify_certs=True,
    connection_class=RequestsHttpConnection,
    timeout=30,
    max_retries=10,
    retry_on_timeout=True,
)


def create_vectorized_index():
    index = "vector-test"
    body = {
        "mappings": {
            "properties": {
                "productInfo": {
                    "type": "text"
                },
                "vector": {
                    "type": "knn_vector",
                    "dimension": 1536
                }
            }
        }
    }

    es.indices.create(index, body=body)


def vectorize(filepath):
    model = InceptionResNetV2()
    layer_name = 'avg_pool'
    intermediate_layer_model = Model(inputs=model.input, outputs=model.get_layer(layer_name).output)

    img = image.load_img(filepath, target_size=(299, 299))
    x = image.img_to_array(img)
    x = np.expand_dims(x, axis=0)
    x = preprocess_input(x)
    intermediate_output = intermediate_layer_model.predict(x)
    return intermediate_output[0]


# print(vectorize("./crawling_images/crawling_images/20221210_13189_어몽어스 짤/001.jpg"))


def index_to_vector(info, vector):
    index = "vector-test"
    doc = {
        'productInfo': info,
        'vector': vector
    }

    response = es.index(index=index, body=doc)
    return response


def search_for_vector(vector):
    index = "vector-test"
    body = {
        "size": 20,
        "query": {
            "knn": {
                "vector": {
                    "vector": vector,
                    "k": 20
                }
            }
        }
    }

    response = es.search(index=index, body=body)

    return response['hits']['hits']


if __name__ == "__main__":
    # print(es.info())

    # print(create_vectorized_index())

    dir_path = "./tagged_images/"

    # list to store files
    images = []
    # Iterate directory
    for path in os.listdir(dir_path):
        # check if current path is a file
        if os.path.isfile(os.path.join(dir_path, path)):
            images.append(path)

    # for img in images:
    #     vector = vectorize(dir_path + img)
    #     print(img, vector)
    #     index_to_vector(img, vector)

    # vector = vectorize("./crawling_images/crawling_images/2022125_15556_루피 짤/106.jpg")
    # vector = vectorize("./crawling_images/crawling_images/2022125_14551_페페 짤/015.jpg")
    vector = vectorize("./img/똘페2.jpeg")
    # vector = vectorize("./crawling_images/crawling_images/20221210_13189_어몽어스 짤/018.jpg")
    result = search_for_vector(vector)

    for data in result:
        # print(data['_source']['productInfo'])
        print(data['_score'], data['_source']['productInfo'])