Visual Moderation

Sieve's image moderation provides a robust solution for moderating images in real-time, designed to identify inappropriate or unwanted content across various hierarchical categories. The app enables users to moderate at different levels, providing flexibility and granular control over content moderation.

Quick Links

• Pricing
• Moderation Levels
• Response Format
• Batch Processing

Key Features

• Hierarchical Moderation: Leverages a three-level classification system, allowing for both broad and specific content targeting.
• Customizable Moderation: Includes a whitelist feature to exclude specific classes from detection, allowing tailored moderation to suit specific needs.
• Detailed Classifications: Provides specific subcategories for precise content identification, such as distinguishing between different types of nudity or violence.
• Adjustable Frame Rate: Developers can set the frame rate for video moderation, enabling flexible processing. For example, setting the frame rate to 0.5 fps would analyze one frame every two seconds.
• Batch Processing Support: Pass .zip or .tar files containing multiple images and videos to process them all at once, for a more cost-effective solution.

Pricing

Processing Type	Cost / 1k Images	Cost / Min Video
Standard Processing	$0.95	$0.095*
With Celebrity Detection	$1.25	$0.125*
Batch Processing	$0.35	$0.095*
Batch Processing with Celebrity Detection	$0.58	$0.125*

* Assuming 1 fps

Note: Batch processing rates apply to groups of 5 images in a .zip file. For example, if processing 10 images, batch processing prices will be applicable to all images. If processing 6 images, 5 images will use batch pricing while 1 image will use standard pricing.

Moderation Levels

The app uses a three-level classification system:

1. Level-1: Broad categories (e.g., Explicit, Violence, Drugs & Tobacco)
2. Level-2: More specific subcategories (e.g., Explicit Nudity, Graphic Violence, Products)
3. Level-3: Highly detailed classifications (e.g., Exposed Male Genitalia, Weapon Violence, Pills)

Note: Developers have the flexibility to moderate content at any level or combination of levels:

1. Level 1 (Broad): Allows for high-level content filtering. For example, a user can block all Explicit content.

2. Level 2 (Specific): Enables more granular control. For instance, a user might allow images of "Alcoholic Beverages" while blocking "Drugs & Tobacco Paraphernalia & Use".

3. Level 3 (Highly Specific): Offers the most precise content moderation. A user could, for example, permit images of "Drinking" but prohibit those depicting "Smoking".

This multi-tiered approach allows for customized and nuanced content moderation strategies tailored to specific needs and preferences.

Moderation Classes

Level-1	Level-2	Level-3	Definitions
Explicit	Explicit Nudity	Exposed Male Genitalia	Human male genitalia, including the penis (whether erect or flaccid), the scrotum, and any discernible pubic hair.
		Exposed Female Genitalia	External parts of the female reproductive system, encompassing the vulva, vagina, and any observable pubic hair.
		Exposed Buttocks or Anus	Human buttocks or anus, including instances where the buttocks are nude or when they are discernible through sheer clothing.
		Exposed Female Nipple	Human female nipples, including fully visible and partially visible areola (area surrounding the nipple) and nipples.
	Explicit Sexual Activity	N/A	Depiction of actual or simulated sexual acts which encompasses human sexual intercourse, oral sex, as well as male genital stimulation and female genital stimulation by other body parts and objects.
	Sex Toys	N/A	Objects or devices used for sexual stimulation or pleasure, e.g., dildo, vibrator, butt plug, beads, etc.
Non-Explicit Nudity of Intimate parts and Kissing	Non-Explicit Nudity	Bare Back	Human posterior part where the majority of the skin is visible from the neck to the end of the spine.
		Exposed Male Nipple	Human male nipples, including partially visible nipples.
		Partially Exposed Buttocks	Partially exposed human buttocks. This includes a partially visible region of the buttocks or butt cheeks due to short clothes, or partially visible top portion of the anal cleft.
		Partially Exposed Female Breast	Partially exposed human female breast where one a portion of the female's breast is visible or uncovered while not revealing the entire breast.
		Implied Nudity	An individual who is nude, either topless or bottomless, but with intimate parts such as buttocks, nipples, or genitalia covered, occluded, or not fully visible.
	Obstructed Intimate Parts	Obstructed Female Nipple	Visual depiction of a situation in which a female's nipples is covered by opaque clothing or coverings, but their shapes are clearly visible.
		Obstructed Male Genitalia	Visual depiction of a situation in which a male's genitalia or penis is covered by opaque clothing or coverings, but its shape is clearly visible.
	Kissing on the Lips	N/A	Depiction of one person's lips making contact with another person's lips.
Swimwear or Underwear	Female Swimwear or Underwear	N/A	Human clothing for female swimwear (e.g., one-piece swimsuits, bikinis, tankinis, etc.) and female underwear (e.g., bras, panties, briefs, lingerie, thongs, etc.)
	Male Swimwear or Underwear	N/A	Human clothing for male swimwear (e.g., swim trunks, boardshorts, swim briefs, etc.) and male underwear (e.g., briefs, boxers, etc.)
Violence	Weapons	N/A	Instruments or devices used to cause harm or damage to living beings, structures, or systems. This includes firearms, sharp weapons, explosives and ammunition.
	Graphic Violence	Weapon Violence	The use of weapons to cause harm, damage, injury, or death to oneself, other individuals, or properties.
		Physical Violence	The act of causing harm to other individuals or property (e.g., hitting, fighting, pulling hair, etc.) or other act of violence involving crowd or multiple individuals.
		Self-Harm	The act of causing harm to oneself, often by cutting body parts such as arms or legs, where cuts are typically visible.
		Blood & Gore	Visual representation of violence on a person, a group of individuals, or animals, involving open wounds, bloodshed, and mutilated body parts.
		Explosions and Blasts	Depiction of a violent and destructive burst of intense flames with thick smoke or dust and smoke erupting from the ground.
Visually Disturbing	Death and Emaciation	Emaciated Bodies	Human bodies that are extremely thin and undernourished with severe physical wasting and depletion of muscle and fat tissue.
		Corpses	Human corpses in the form of mutilated bodies, hanging corpses, or skeletons.
	Crashes	Air Crash	Incidents of air vehicles, such as airplanes, helicopters, or other flying vehicles, resulting in damage, injury, or death. This term applies when parts of the air vehicles are visible.
Drugs & Tobacco	Products	Pills	Small, solid, often round or oval-shaped tablets or capsules. This term applies to pills presented as standalone, in a bottle, or a transparent packet and does not apply to a visual depiction of a person taking pills.
	Drugs & Tobacco Paraphernalia & Use	Smoking	The act of inhaling, exhaling, and lighting up burning substances including cigarettes, cigars, e-cigarettes, hookah, or joint.
Alcohol	Alcohol Use	Drinking	The act of drinking alcoholic beverages from bottles or glasses of alcohol or liquor.
	Alcoholic Beverages	N/A	Close up of one or multiple bottles of alcohol or liquor, glasses or mugs with alcohol or liquor, and glasses or mugs with alcohol or liquor held by an individual.
Rude Gestures	Middle Finger	N/A	Visual depiction of a hand gesture with middle finger is extended upward while the other fingers are folded down.
Gambling	N/A	N/A	The act of participating in games of chance for a chance to win a prize in casinos, e.g., playing cards, blackjacks, roulette, slot machines at casinos, etc.
Hate Symbols	Nazi Party	N/A	Visual depiction of symbols, flags, or gestures associated with Nazi Party.
	White Supremacy	N/A	Visual depiction of symbols or clothings associated with Ku Klux Klan (KKK) and images with confederate flags.
	Extremist	N/A	Images containing extremist and terrorist group flags.
Children Under 18	N/A	N/A	Images containing children under the age of 18.
Copyright	Copyright Movie or Television Shows	N/A	Depiction of frames from famous movies or TV shows. Only Flag if you are 100pc sure about the movie or TV show. Youtube videos are not covered by this flag.
	Sports Activity	N/A	Depiction of any sports-related activity, including but not limited to team sports, individual sports, athletic activities, training, or recreational sports activities.
Virtual Flags	Video Game Violence	N/A	Depiction of Murder, Killing in a video game, If there is real life violence this flag doesnt apply.
	Video Game Weapons	N/A	Guns, Knives, Explosives, or Bombs in video games only, if there is real life violence this flag doesnt apply.

Celebrity Detection (Alpha)

The Visual Moderation API includes an alpha feature for celebrity detection that can identify public figures in images and videos. The detection currently supports celebrities in the following categories:

• Actors/Actresses
• Musicians
• Athletes
• Politicians
• Other public figures

This is an experimental feature and accuracy may vary. To enable celebrity detection, set celebrity_detection=True when calling the API.

Response Format

The API returns an array of results, with fields varying based on the input type (image, video, or archive file).

Response Fields

• moderation_results: Array of moderation results
  • labels: Array of detected content labels (each with L1, L2, L3 classifications)
  • start & end: (Video only) Timestamps in seconds showing when content was detected
• celebrity_detection: (Optional) Array of celebrity detection results, present when celebrity_detection=True
  • For videos: Contains segments with start, end, and labels array
  • For images: Direct array of detected celebrities
• file_path: (Archive files only) Path of the file within the archive

These are the fields that will be present in the response, depending on the input type and the flags used:

[
  {
    "moderation_results": [
      {
        "start": <start time in seconds>, // Only present if video
        "end": <end time in seconds>, // Only present if video
        "labels": [  // Always present, may be empty array
          {
            "L1": "<label of level 1>",
            "L2": "<label of level 2>",
            "L3": "<label of level 3>"
          }
        ]
      }
    ],
    "celebrity_detection": [  // Only present if celebrity_detection=True
      {
        "celebrity_name": <name of celebrity>,
        "category": <category of celebrity>
      }
    ],
    "file_path": <path of file within archive>  // Only present if archive file
  }
]

Batch Processing

The Visual Moderation API supports batch processing through archive files (.zip, .tar, .tar.gz, .tgz) containing multiple images and/or videos. This provides a more cost-effective way to process multiple files at once.

Supported Archive Types

• .zip
• .tar
• .tar.gz
• .tgz

Note: The archive must only contain images and videos. Any other files present are ignored.

How Batch Processing Works

Batch processing optimizes performance by processing images in parallel groups of 5. For example, 13 images would be split into two batches of 5 and 3 images processed individually, while maintaining the original order. This approach improves throughput through parallel processing and offers cost benefits, with batch pricing applied to complete groups of 5 images while any remainder uses standard pricing.

Response Format

The API returns results that maintain the original file structure from the archive. Each result includes a file_path showing the location within the archive, along with standard moderation results.

Example response for a batch containing one video and one image:

[
  {
    "moderation_results": [
      {
        "start": 0,
        "end": 10,
        "labels": []
      }
    ],
    "celebrity_detection": [
      {
        "start": 0,
        "end": 10,
        "labels": [
          {
            "celebrity_name": "John Doe",
            "category": "athlete"
          }
        ]
      }
    ],
    "file_path": "test_video.mp4"
  },
  {
    "moderation_results": [
      {
        "labels": []
      }
    ],
    "celebrity_detection": [],
    "file_path": "test_image.png"
  }
]