This post is using machine translation. Sorry if it is difficult to understand.
I am using FACE DETECTION and DOCUMENT TEXT DETECTION.
I am trying to retrieve face and text information from an image.
I am sending the same request using cURL but sometimes I get different responses back.
Am I wrong in thinking that identical requests will get identical responses?
Is there a way to get an identical response every time?
The following is an example of FACE_DETECTION.
- input.json
{
"requests": [
{
"image": {
"content": "/9j......",
},
"features": [
{
"type": "FACE_DETECTION",
"model": "builtin/stable"
}
],
"imageContext": {
"languageHints": ['ja']
}
}
]
}
- request
$ curl -s -X POST -H "Authorization: Bearer "xxxxxxxxxxx -H "Content-Type: application/json; charset=utf-8" -d @input.json https://vision.googleapis.com/v1/images:annotate
Most of the time response 1 is returned, but about 1 out of 10 times response 2 is returned.
- response 1
{
"responses": [
..........
"rollAngle": 0.20178626,
"panAngle": -2.3033228,
"tiltAngle": 0.5400786,
..........
}
]
}
]
}
- response 2
{
"responses": [
..........
"rollAngle": 0.2607577,
"panAngle": -2.2164109,
"tiltAngle": 0.66491675,
..........
}
]
}
]
}