Face detection

This tutorial helps in the detection of human faces in base64 images, using the Halisi-Cow Face Detection API.

To get started, first ensure that you are Signed Up/In on the RapidAPI platform. You may also need to subscribe to Halisi-Animal Face Detection API. Upon successfully signing-up/in and subscribing, you gain automatic access to the API-key, which is essential to unlock more Halisi features! Your RapidAPI key will be automatically displayed in the Keypoints tab, under the X-RapidAPI-Key parameter in the Header Parameters section.

Pre-requisite:

1. Base 64 image.
   If you plan to explore Halisi-Animal API using the RapidAPI platform, please note that the API only accepts base64 input on this platform.
   You can convert the image formats (png, jpg, jpeg, etc.) to base64 using encoding tools or libraries.
   Keep the converted base64 string handy.

Once the pre-requisites are checked, we now proceed to test this API using the following steps:

1. Insert the base64 string in the Request Body section, in the tag- “—Paste your image here—”.
2. Pass values in the range [0 - 1] for Confidence and IoU Thresholds, as optional parameters.
   If you intend to use the default values, proceed directly to step 3.
3. Click the “Test Endpoint” button.
4. A request will be sent to the API alongwith the input parameters. You can see the json response in the adjacent block.

Input:

1. Base64 image
2. Confidence threshold (Optional parameter, default = 0.45)
3. IoU threshold (Optional parameter, default = 0.45)

Output:

1. Bounding box co-ordinates (format: [left_x, top_y, right_x, bottom_y])
2. Detection scores (between 0 and 1)
3. Elapsed time (in miliseconds)
4. Error message

Example request

import requests

url = “https://halisi-animal-face-detection.p.rapidapi.com/DetectCow”

querystring = {“Confidence”:“0.60”,“IoU”:“0.40”}

payload = {“image”: “base64 image pasted here”}

headers = {
“content-type”: “application/json”,
“X-RapidAPI-Key”: “----your RapidAPI key----”,
“X-RapidAPI-Host”: “halisi-animal-face-detection.p.rapidapi.com”
}

response = requests.request(“POST”, url, json=payload, headers=headers, params=querystring)

print(response.text)

Example response

bbox: "[69, 10, 520, 461]"
detection_score: "0.8890097"
elapsed_time: 552
message: “Face successfully detected”