Create AI CM:soft_nudity task

This algorithm allows to identify explicit nudity and partial nudity too (including the presence of male and female faces and other uncovered body parts) in a video. Generic info about all capabilities and limits see in the generic “Content Moderation” method.

What is Soft nudity detection?

This method is often used to analyze UGC to determine whether videos can be published to all users, or to prohibit publication due to offensive and inappropriate content.

Objects that can be detected:

• ANUS_COVERED
• ANUS_EXPOSED
• ARMPITS_COVERED
• ARMPITS_EXPOSED
• BELLY_COVERED
• BELLY_EXPOSED
• BUTTOCKS_COVERED
• BUTTOCKS_EXPOSED
• FACE_FEMALE
• FACE_MALE
• FEET_COVERED
• FEET_EXPOSED
• FEMALE_BREAST_COVERED
• FEMALE_BREAST_EXPOSED
• FEMALE_GENITALIA_COVERED
• FEMALE_GENITALIA_EXPOSED
• MALE_BREAST_EXPOSED
• MALE_GENITALIA_EXPOSED

This method allows you to identify faces and other body parts. Used to find complex combinations of what is happening in a video. Please note that the number of objects is more than in the hard-nudity-detection. The method is slower.

How to use?

The information is returned with the video frame number where it was found and probability of the detected object. Nudity detection is done using AI, so for each object a probability percentage is applied; objects with a probability of at least 30% are included in the response.

Video processing speed is approximately 1:5.

Example of detected nudity or body parts:

{
  "nudity_detected": true,
  "detection_results": [ "FACE_FEMALE", "BELLY_COVERED" ]
  "frames": [
      {
          "confidence": 0.82,
          "frame_number": 1,
          "label": "BELLY_COVERED"
      },...
  ]
}

Example response when nudity or body parts were not found:

{
  "nudity_detected": false,
  "detection_results": []
  "frames": []
}

There is no universal recipe under which a video can be considered unacceptable, since different services host different types of videos for different audiences: adult content, children’s content, educational content, etc. You can determine the probability threshold at which you consider a video inappropriate. The easiest option is to run several of your videos and analyze the resulting probability coefficient.

Sometimes a detected object at the beginning of the video immediately makes it clear that there is no need to further analyze the video. For such cases, you can use stop tags. Use parameter “stop_objects” to specify comma separated stop tags. It is also possible to specify % probability threshold value, above which the stop tag will be triggered.

  {
      "url": "...",
      "stop_objects": "BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED"
  }

Please note that the API only provides a set of data (json) about the objects found, so no video is generated. The demo video video (above ^) was specially created based on json from the API for visual demonstration and better perception of the possibilities.

POST /streaming/ai/tasks#cm_soft_nudity

Create AI CM:soft_nudity task

cURL

curl --request POST \
  --url 'https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity' \
  --header 'Authorization: <api-key>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "url": "https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4",
  "task_name": "content-moderation",
  "category": "soft_nudity",
  "stop_objects": "BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED"
}
'

import requests

url = "https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity"

payload = {
    "url": "https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4",
    "task_name": "content-moderation",
    "category": "soft_nudity",
    "stop_objects": "BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED"
}
headers = {
    "Authorization": "<api-key>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    url: 'https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4',
    task_name: 'content-moderation',
    category: 'soft_nudity',
    stop_objects: 'BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED'
  })
};

fetch('https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'url' => 'https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4',
    'task_name' => 'content-moderation',
    'category' => 'soft_nudity',
    'stop_objects' => 'BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED'
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: <api-key>",
    "Content-Type: application/json"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity"

	payload := strings.NewReader("{\n  \"url\": \"https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4\",\n  \"task_name\": \"content-moderation\",\n  \"category\": \"soft_nudity\",\n  \"stop_objects\": \"BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Authorization", "<api-key>")
	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity")
  .header("Authorization", "<api-key>")
  .header("Content-Type", "application/json")
  .body("{\n  \"url\": \"https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4\",\n  \"task_name\": \"content-moderation\",\n  \"category\": \"soft_nudity\",\n  \"stop_objects\": \"BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED\"\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.gcore.com/streaming/ai/tasks#cm_soft_nudity")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Authorization"] = '<api-key>'
request["Content-Type"] = 'application/json'
request.body = "{\n  \"url\": \"https://demo-files.gvideo.io/ai_demo_subtitles_nudity_detection.mp4\",\n  \"task_name\": \"content-moderation\",\n  \"category\": \"soft_nudity\",\n  \"stop_objects\": \"BELLY_COVERED:0.9,FEMALE_GENITALIA_COVERED\"\n}"

response = http.request(request)
puts response.read_body

201

{
  "task_id": "aafe70c6-0000-0000-0000-327b65f7670f"
}

{
  "error": "Queue limit reached (100), try later."
}

{
  "error": "Feature is disabled. Contact support to enable."
}

Authorizations

Authorization · string · header · required

API key for authentication. Make sure to include the word apikey, followed by a single space and then your token. Example: apikey 1234$abcdef

Body

application/json

task_name · enum<string> · required

Name of the task to be performed

Available options: content-moderation

url · string · required

URL to the MP4 file to analyse. File must be publicly accessible via HTTP/HTTPS.

category · enum<string> · required

AI content moderation with "soft_nudity" algorithm

Available options: soft_nudity

client_user_id · string

Meta parameter, designed to store your own identifier. Can be used by you to tag requests from different end-users. It is not used in any way in video processing.

Maximum string length: 256

client_entity_data · string

Meta parameter, designed to store your own extra information about a video entity: video source, video id, etc. It is not used in any way in video processing.

For example, if an AI-task was created automatically when you uploaded a video with the AI auto-processing option (nudity detection, etc), then the ID of the associated video for which the task was performed will be explicitly indicated here.

Maximum string length: 4096

stop_objects · enum<string>

Comma separated objects, and probabilities, that will cause the processing to stop immediatelly after finding.

Available options: ANUS_COVERED, ANUS_EXPOSED, ARMPITS_COVERED, ARMPITS_EXPOSED, BELLY_COVERED, BELLY_EXPOSED, BUTTOCKS_COVERED, BUTTOCKS_EXPOSED, FACE_FEMALE, FACE_MALE, FEET_COVERED, FEET_EXPOSED, FEMALE_BREAST_COVERED, FEMALE_BREAST_EXPOSED, FEMALE_GENITALIA_COVERED, FEMALE_GENITALIA_EXPOSED, MALE_BREAST_EXPOSED, MALE_GENITALIA_EXPOSED

Response

201

application/json

Response returns ID of the created AI task. Using this AI task ID, you can check the status and get the video processing result. Look at GET /ai/results method.

task_id · string<uuid> · required

ID of the created AI task, from which you can get the execution result