Deep learning

Realtime Image and Text Moderation using AI

[ 2 ] ​Make call to Image Moderation API

Use this method to submit a photo URL to the Automated Intelligent Moderation. Using industry standard Machine Learning models the service will return a probability likelihood based on (0 – 100) percentage of class categories that the submitted photo contains.

Images that are at least 300px by 300px will return the most accurate results, however the minimum image size accepted is 50px by 50px.

IMAGE MAX SIZE OF 5 Mb WILL BE SCANNED

End point to generate access token (REST API - POST):
https://checkmycontent.net/image


Request body parameter: img    [your valid image url that is publicly accessible]

Example:

curl -X POST \ 'https://checkmycontent.net/image' \ -H 'authorization: BEARER Insert-Your-Access-Token-Here' \ -H 'cache-control: no-cache' \ -H 'content-type: application/json' \ -d '{ "img" : "https://picsum.photos/1920/1080" }'

POST Response:

200 OK: { result : { "predictions" : [ { "class_name" : "00.0000" }, ... {} , {} ] , "classifications" : [ { type: "predicted object in image", probability: "00.0000"} ... {} , {} ] }
400: Error: { msg : "E001: Access token missing." }
400: Error: { msg : "E002: Authorization Error: Your token has expired!" }
400: Error: { msg : "E003: Authorization Error: Your token is malformed!" }
400: Error: { msg : "E004: img input parameter not provided." }
400: Error: { msg : "E005: Image format not supported. Convert to .jpg, .jpeg, .webp or .png format and retry." }
400: Error: { msg : "E006: Image size greater than 5mb. Convert to .jpg, .jpeg., .webp or .png format and reduce the size to less than 5mb." }
400: Error: { msg : "E500: Generic Error: (err.msg)" }

[ 3 ] ​Make call to Text Moderation API

Use this method to submit a text string to the Automated Intelligent Moderation. Using industry standard Machine Learning models the service will return a percentage probability (0 – 100) of class categories that the submitted text contains.

The analysis by the AI model is close to 89% accurancy.

MAX 500 WORDS PER API CALL AND ENGLISH ONLY

End point to generate prediction for the text (REST API - POST):
https://checkmycontent.net/text

Request body parameter: text [max 500 words]

Example:

curl -X POST \ 'https://checkmycontent.net/text' \ -H 'authorization: BEARER Insert-Your-Access-Token-Here' \ -H 'cache-control: no-cache' \ -H 'content-type: application/json' \ -d '{ "text" : "Your text goes here" }'

POST Response:

200 OK: { result : { "predictions" : [ { "class_name" : "00.0000" }, ... {} , {} ] , "classifications" : [ { type: "predicted object in image", probability: "00.0000"} ... {} , {} ] }
400: Error: { msg : "E001: Access token missing." }
400: Error: { msg : "E002: Authorization Error: Your token has expired!" }
400: Error: { msg : "E003: Authorization Error: Your token is malformed!" }
400: Error: { msg : "E004: text input parameter not provided." }
400: Error: { msg : "E500: Generic Error: (err.msg)" }