Hive ModelsModeration Dashboard
Docs
Log inSign up

Visual Moderation - Overview

Introduction (V2)

Note: New V3 API

Hive's V3 Visual Moderation API is open for self-serve use—no project setup with Sales required.

  • Test instantly in the Playground (100 req/day limit).
  • View the V3 API Docs and cURL samples shown in the Playground to integrate at the default rate limits.
  • Need higher throughput? Reach out to us by filling out our contact form to enable an Enterprise project with higher limits.

Hive's Visual Moderation classifies an entire image or video into different categories and assigns a confidence score to each class.

Visual Moderation is multi-headed, where groups of mutually exclusive model classes belong to a single model head. For example, when an image is run through Hive's visual moderation model, one head might classify sexually not-safe-for-work (NSFW) content while another head might classify the presence of guns.

This concept is illustrated below. This imaginary model has two heads:

NSFW classification: general_nsfw, general_suggestive, general_not_nsfw_not_suggestive

Gun classification: gun_in_hand, animated_gun, gun_not_in_hand, no_gun

The confidence scores for each model head sum to 1.

{
  "output": [
    {
      "time": 0,
      "classes": [
        {
          "class": "general_nsfw",
          "score": 0.9
        },
        {
          "class": "general_suggestive",
          "score": 0.05
        },
        {
          "class": "general_not_nsfw_not_suggestive",
          "score": 0.05
        },
        {
          "class": "gun_in_hand",
          "score": 0.88
        },
        {
          "class": "animated_gun",
          "score": 0.04
        },
        {
          "class": "gun_not_in_hand",
          "score": 0.04
        },
        {
          "class": "no_gun",
          "score": 0.04
        }
      ]
    }
  ]
}

When submitting a video to be processed, Hive’s backend splits the video into frames, runs the model on each frame, then recombines the results into an aggregated response for the entire video. The video output for a classifier is similar to a list of classification output objects, but with multiple timestamps.

Livestream Support: Hive also supports processing live RTMP and HLS streams for Visual Moderation. Customers submit a live stream URL, and we sample said stream and submit the given frames to their project. Alternatively, customers can capture short clips (less than or equal to 30 seconds) and send those as tasks to both our visual and audio moderation APIs.

Our stream submission guide details how to submit a live stream to our endpoint, which supports generating image frames out of a video stream. By default, frames are generated at a rate of 1 frame per second. Note: You must contact Hive to have streams allocated to your project before you can use this feature.

A more detailed walkthrough on how to submit visual classification tasks via the API and how to interpret the visual model response can be found in our customer guide. To try out our model, visit our demo hub where you can submit your own images and get a sense of what our results look like.

Visual Moderation Classes

Hive's visual classification models support a wide variety of classes that are relevant to content moderation. Broadly, visual moderation classes can be separated into five main categories: sexual content, violent imagery, drugs, hate imagery, and image attributes. When deciding how to process our API response in order to implement your content policy, you should consult the following class descriptions to decide which classes to moderate.

📘

NOTE:

This page simply lists supported visual moderation classes and gives a brief description. For a more detailed breakdown of subject matter flagged by each class, click the class name or go to the Class Descriptions subpages.

Note: Older versions of the API might not perfectly match the outline below. Please reach out to [email protected] if you would like to access the latest content moderation classes.

Sexual

NSFW Head:

  • general_nsfw - genitalia, sexual activity, nudity, buttocks, sex toys, animal genitalia
  • general_suggestive - shirtless men, underwear / swimwear, sexually suggestive poses without genitalia, occluded or blurred sexual activity
  • general_not_nsfw_not_suggestive - none of the above, clean

Sexual Activity Head:

  • yes_sexual_activity - a sex act or stimulation of genitals are present in the scene
  • no_sexual_activity - no sex act is present in the scene

Realistic NSFW Head:

  • yes_realistic_nsfw - live nudity, sex acts, or photo-realistic representations of nudity or sex acts
  • no_realistic_nsfw - non-photorealistic representations of nudity or sex acts (statues, crude drawings, paintings etc.); lack of any NSFW content

Female Underwear Head:

  • yes_female_underwear - lingerie, bras, panties
  • no_female_underwear

Bra Head:

  • yes_bra - standard bras, including ones worn under sheer clothing
  • no_bra

Panties Head:

  • yes_panties - women's underwear, including boyshorts and thongs
  • no_panties

Negligee Head:

  • yes_negligee - negligee, chemises, and other sheer, nightgown-type garments
  • no_negligee

Male Underwear Head:

  • yes_male_underwear - fruit-of-the-loom, boxers
  • no_male_underwear

Sex Toy Head:

  • yes_sex_toy - dildos, certain lingerie
  • no_sex_toy

Cleavage Head:

  • yes_cleavage - identifiable female cleavage
  • no_cleavage

Female Nudity Head:

  • yes_female_nudity - breasts or female genitalia
  • no_female_nudity

Male Nudity Head:

  • yes_male_nudity - male genitalia
  • no_male_nudity

Female Swimwear Head:

  • yes_female_swimwear - bikinis, one-pieces, not underwear
  • no_female_swimwear

Bodysuits Head:

  • yes_bodysuit - bodysuits that do not cover the thigh, including corsets
  • no_bodysuit

Miniskirts Head:

  • yes_miniskirt - skirts that end above the mid-thigh
  • no_miniskirt

Sports Bra Head:

  • yes_sports_bra - sports bras, bralettes, bandeaus, and other bra-like clothing
  • no_sports_bra

Sportswear Bottoms Head:

  • yes_sportswear_bottoms - bottoms worn during exercise
  • no_sportswear_bottoms

Bulges Head (Beta)

  • yes_bulge - penises that are visible underneath clothing
  • no_bulge

Breast Head (Beta)

  • yes_breast - female breasts where nipples or areolas are visible
  • no_breast

Genitals Head (Beta)

  • yes_genitals - human genitals (vulvas, penises, testicles)
  • no_genitals

Butt Head (Beta)

  • yes_butt - exposed butts
  • no_butt

Tongue Head (Beta)

  • kissing - mouth-to-mouth contact as well as cheek and forehead kisses
  • licking - mouth-to-body contact of any kind, including oral sex
  • no_tongue

Shirtless Male Head:

  • yes_male_shirtless - shirtless below mid-chest
  • no_male_shirtless

Sexual Intent Head:

  • yes_sexual_intent - occluded, blurred, or hidden sexual activity
  • no_sexual_intent

Undressed Head:

  • yes_undressed - a subject is nude/unclothed, even if genitals etc. are not visible due to pose or digital overlay, or are covered by an object
  • no_undressed - underwear, swimwear, shirtless men if not evident they are nude

Animal Genitalia Head:

  • animal_genitalia_and_human - sexual activity including both animals and humans
  • animal_genitalia_only - animals mating and pictures of animal genitalia
  • animated_animal_genitalia - drawings of sexual activity involving animals
  • no_animal_genitalia - none of the above, clean

Violence

Gun Head:

  • gun_in_hand - person holding rifle, handgun
  • gun_not_in_hand - rifle, handgun, not in hand
  • animated_gun - gun in games, cartoons, etc. can be in-hand or not.
  • no_gun

Knife Head:

  • knife_in_hand - person holding knife, sword, machete, razor blade
  • knife_not_in_hand - knife, sword, machete, razor blade, not in hand (outside of culinary settings)
  • culinary_knife_not_in_hand - culinary knives not being held or handled by a person
  • culinary_knife_in_hand - knife being used for preparing food
  • no_knife

Blood Head:

  • very_bloody - gore, visible bleeding, self-cutting
  • a_little_bloody - fresh cuts / scrapes, light bleeding
  • other_blood - animated blood, fake blood, animal blood such as game dressing
  • no_blood - minor scabs, scars, acne, etc. are not considered ‘blood’ by model

Hanging Head:

  • hanging - the presence of a human hanging by noose (dead or alive)
  • noose - a noose is present in the image with no human hanging from it
  • no_hanging_no_noose - no person hanging and no noose present

Corpses Head (Beta):

  • human_corpse - human dead body present in image
  • animated_corpse: animated dead body present in image
  • no_corpse

Emaciated Bodies Head:

  • yes_emaciated_body - emaciated human or animal body present in image
  • no_emaciated_body

Self Harm Head:

  • yes_self_harm - self cutting, burning, instances of suicide or other self harm methods present in image
  • no_self_harm

Animal Abuse Head (Beta)

  • yes_animal_abuse - animals being beaten, tortured, or treated inhumanely as well as animals with graphic injuries
  • no_animal_abuse

Fights Head (Beta)

  • yes_fight - two or more people engaging in a physical fight
  • no_fight

Child Safety Head (Beta)

  • yes_child_safety - shirtless child 11 years old or younger present in the image
  • no_child_safety

Drugs and other vices

Pill Head:

  • yes_pills - pills and / or drug powders
  • no_pills - no pills and / or drug powders

Injectable Head:

  • illicit_injectables - heroin and other illegal injectables
  • medical_injectables - injectables for medical use
  • no_injectables - no injectable drug paraphernalia

Smoking Head:

  • yes_smoking - cigarettes, cigars, marijuana, vapes, or other smoking paraphernalia
  • no_smoking - no cigars, marijuana, vapes, or other smoking paraphernalia

Marijuana Head (Beta)

  • yes_marijuana - marijuana or marijuana-related paraphernalia
  • no_marijuana

Gambling Head:

  • yes_gambling - depicts gambling activity like slot machines, casino games, sports betting, or lottery where betting is visible or implied
  • no_gambling - no gambling activity, regular card/dice games or competitive games with no evidence of betting

Alcohol Head:

  • yes_drinking_alcohol - depicts alcoholic beverages being consumed
  • yes_alcohol - depicts alcoholic beverages, present but not being consumed
  • animated_alcohol - depicts alcoholic beverages in animated movies, cartoons, or art
  • no_alcohol - does not depict alcoholic beverages or identifiable alcohol use

Hate

Nazi Head:

  • yes_nazi - Nazi symbols
  • no_nazi - absence of the above

Terrorist Head:

  • yes_terrorist - ISIS flag
  • no_terrorist - absence of the above

KKK Head:

  • yes_kkk - KKK symbols
  • no_kkk - absence of the above

Confederate Flag Head:

  • yes_confederate - shows the Confederate "stars and bars," including graphics, clothing, tattoos, and spinoff flags
  • no_confederate - absence of the above

Middle Finger Head:

  • yes_middle_finger - middle finger
  • no_middle_finger - absence of the above

Other Attributes

Text Head:

  • text - any form of text or writing is present somewhere on the image
  • no_text - no text present in the image

Overlay Text Head:

  • yes_overlay_text - digitally overlaid text is present on an image (think meme text)
  • no_overlay_text - lack of digitally overlaid text in the image

Child Presence:

  • yes_child_present: a baby or toddler is present in the image
  • no_child_present

Religious Icons (Beta)

  • yes_religious_icon: a religious icon is present in this image
  • no_religious_icon

Drawings (Beta):

  • yes_drawing: a drawing, painting, or sketch is the central part of the image
  • no_drawing

Image Type Head:

  • animated - the image is animated
  • hybrid - the image is partially animated
  • natural - the image has no animation

QR Codes Head:

  • yes_qr_code - the image contains a QR code
  • no_qr_code - the image does not contain a QR code

Brand Safety & Suitability - GARM taxonomy

Hive's Brand Safety and Brand Suitability APIs are powered by Hive's visual moderation model and are additionally mapped to the GARM Brand Safety & Suitability Framework (Global Alliance for Responsible Media), which was established as an industry-standard for categorizing harmful content. For more information click here.

Request Format (V2)

The request format for this API includes a field for the media being submitted, either as a local file path or as a url. For more information about submitting a task, see our API reference guides to synchronous and asynchronous submissions.

# Submit a task with an image URL
curl --request POST \
  --url https://api.thehive.ai/api/v2/task/sync \ # this is a sync example, see API reference for async
  --header 'accept: application/json' \
  --header 'authorization: token <API_KEY>' \
  --form 'url=http://hive-public.s3.amazonaws.com/demo_request/gun1.jpg'

# Submit a task with a local file
 curl --request POST \
     --url https://api.thehive.ai/api/v2/task/sync \ # this is a sync example, see API reference for async
     --header 'Authorization: Token <token>' \
     --form 'media=@"<absolute/path/to/file>"'

To get started with V3, check out our Playground and go to our "API" tab for code samples.


Choosing Thresholds for Visual Moderation

Note: V3 returns the same class names and confidence scores as V2, so your threshold logic stays unchanged.

For each of the classes mentioned above, you will need to set thresholds to decide when to take action based on our model results. For optimum results, a proper threshold analysis on a natural distribution of your data is recommended (for more on this please contact Hive at the email below). Generally, though, a model confidence score threshold of >.90 is a good place to start to flag an image for any class of interest.

For questions on best practices, please message your point of contact at Hive or send a message to [email protected] to contact our API team directly.

Supported File Types

Image Formats:
gif
jpg
png
webp

Video Formats:
mp4
webm
avi
mkv
wmv
mov

Updated about 1 month ago