Livestream

6 min

Nima allows you to automatically detect harmful content in livestreams using the AI detection providers you have configured.

Example payload

{
    "priority": "high",
    "channel": "automated detection",
    "reason_for_request": "This video is a live stream",
    "note": "This video is a live stream",
    "customerSpecific": {},
    "content": {
        "content_id": "some_content_id_xxx",
        "title": "title_for_live_stream",
        "body": "body_for_live_stream",
        "url": "https://zssd-kijami.hls.camzonecdn.com/CamzoneStreams/zssd-kijami/Playlist.m3u8",
        "language": "en",
        "token": "",
        "type": {
            "id": "VIDEO_STREAM",
            "name": "Video Stream"
        },
        "status": "published",
        "customerSpecific": {
        
        }
    },
    "reportee": {
        "id": "reportee_id",
        "name": "reportee_name",
        "phoneNumber": "reportee_phone_numner",
        "emailAddress": "reportee_email_address",
        "customerSpecific": {}
    },
    "location": {
        "city": "location_city",
        "postalCode": "location_postal_code",
        "region": "location_region",
        "countryCode": "location_country_code",
        "ipAddress": "location_id_address",
        "customerSpecific": {}
    },
    "label": [
        "stream"
    ]
}

Request payload requirements

For livestream moderation, ensure your request payload includes the following content-type specific requirements:

Attribute	Required	Description
content.url	Yes	The URL should point to the livestream using a supported protocol
content.type.id	Yes	Must be set to "VIDEO_STREAM"
content.token	No	Optional token for livestream authentication

Configuring detection rules

To create a rule for automated livestream detection, navigate to Configure > Trust & Safety > Rules and define the following conditions:

Required Conditions:

Channel: Must be set to "Proactive Detection"

Content Type: Must be set to "Video Stream"

Once these conditions are set, the list of available AI detection providers will be accessible via the Choose Provider button.

Livestream-specific features

How it works

Your backend submits a livestream to Nima API

The livestream analyzer monitors the stream in real-time and creates alerts each time unwanted content is detected

Moderation results are sent to Nima as they occur

Available detection models

The following detection models are commonly used for livestream moderation (for a full list, please contact support):

Model	Description
nudity-2.1	Detect all types of nudity and sexual content. Pornography, X-rated nudity, partial nudity, suggestive scenes and poses, lingerie... More
violence	Detect physical violence, threats and more
offensive	Detect offensive and hateful signs, symbols, flags and gestures.
weapon	Detect weapons, firearms and threatening knives
face-attributes	Detect faces along with their main attributes: gender, age...
gore-2.0	Detect horrific imagery including scenes with blood, wounds, self-harm, guts...
recreational_drug	Detect recreational and medical drugs
alcohol	Detect wine, beer, cocktails, spirits.
gambling	Detect gambling, casinos, slot machines...
tobacco	Detect smoking and tobacco products
money	Detect displays of money and banknotes
self-harm	Detect self-harm

Supported protocols

The livestream API supports the following protocols for live video ingestion:

HLS (HTTP Live Streaming) over HTTPS or HTTP

RTMP and RTMPS (Real-Time Messaging Protocol)

RTSP and RTP (Real Time Streaming Protocol)

MPEG-DASH (MPEG Dynamic Adaptive Streaming over HTTP)