Help Center> Content Moderation> Service Overview> Application Scenarios
Updated on 2024-04-22 GMT+08:00
View PDF

Application Scenarios

Image Moderation

Application scenarios are as follows:

  • Live video

    In the webcast scenario, thousands of channels are broadcasting concurrently, making manual review of broadcasting contents impossible. Moderation (Image) monitors, identifies, and flags live video channels with inappropriate, unwanted, or offensive content in real time.

    The advantages are as follows:
    • High accuracy: Moderation (Image) yields incredibly high levels of accuracy in detection with optimized deep learning algorithms.
    • High speed: Moderation (Image) responds to live video requests within 0.1 seconds.
  • Online shopping

    Moderation (Image) identifies and warns of non-compliant images uploaded by sellers and users to prevent such images from being released, reducing manual reviews and non-compliance risks.

    The advantages are as follows:
    • High accuracy: Moderation (Image) yields high levels of accuracy in detection with optimized deep learning algorithms.
    • Rapid response: Moderation (Image) recognizes a single image within 0.1 seconds.
  • Internet forum

    Moderation (Image) detects and flags non-compliant content to help you reduce your legal exposure.

    The advantages are as follows:
    • High accuracy: Moderation (Image) yields incredibly high levels of accuracy in detection with optimized deep learning algorithms.
    • Rapid response: Moderation (Image) recognizes a single image within 0.1 seconds.

Text Moderation

Application scenarios are as follows:

  • E-commerce comment screening

    Moderation (Text) checks product comments on e-commerce websites and identifies non-compliant comments with pornographic elements, spamming, and other types of events to ensure optimal UX.

    The advantages are as follows:

    • High accuracy: Moderation (Text) yields incredibly high levels of accuracy in detection with optimized deep learning algorithms.
    • Rapid response: Moderation (Text) can respond within 0.1 seconds.
  • User nickname review

    Intelligently reviews user registration information on websites and blocks nicknames that contain advertisements and pornographic elements.

    The advantages are as follows:

    • High accuracy: Moderation (Text) yields incredibly high levels of accuracy in detection with optimized deep learning algorithms.
    • Rapid response: Moderation (Text) can respond within 0.1 seconds.
  • Media content review

    Automatically identifies contraband information in media content to avoid non-compliance risks in released articles.

    The advantages are as follows:

    • Fast update: Moderation (Text) continuously and quickly updates the dictionary to identify new non-compliant content in a timely manner.
    • High speed: Moderation (Text) can recognize non-compliant content within 0.1 seconds.
  • Bullet Comment Review

    Moderation (Text) instantly reviews bullet comments that scroll across a screen to ensure webcast quality and reduce non-compliance risks.

    The advantages are as follows:

    • Large-scale dictionary: Moderation (Text) has a large-scale built-in dictionary with support for various matching rules.
    • Fast update: Moderation (Text) continuously and quickly updates the dictionary to identify new non-compliant content in a timely manner.
  • Chat content review

    Moderation (Text) detects potential non-compliant information (such as offensive content, pornographic elements, and reactionary tendencies) in game chats in real time to purify the network environment.

    The advantages are as follows:

    • Large-scale dictionary: Moderation (Text) has a large-scale built-in dictionary with support for various matching rules.
    • Rapid response: Moderation (Text) can respond within 0.1 seconds.

Audio Moderation

Application scenarios are as follows:

  • Online education

    Moderation (Audio) monitors audio teaching content and intelligently reviews violation scenarios such as pornography, violence, abuse, and advertisements.

    The advantages are as follows:
    • High accuracy: Moderation (Audio) uses optimized deep learning algorithms to improve the voice moderation accuracy in complex environments.
    • Recognition of special sounds: Moderation (Audio) provides models for recognizing special sounds, such as copulation calls, moaning, and sensitive voiceprints.
  • Games/Social media apps

    Moderation (Audio) monitors the chat content and voice feeds in games and social media apps to reduce non-compliance risks.

    The advantages are as follows:

    • High accuracy: Moderation (Audio) uses optimized deep learning algorithms to improve the voice moderation accuracy in complex environments.
    • Recognition of special sounds: Moderation (Audio) provides models for recognizing special sounds, such as copulation calls, moaning, and sensitive voiceprints.
  • Recording/Radio

    Audio Moderation monitors the audio data of content transmission and FM radio to reduce non-compliance risks.

    The advantages are as follows:
    • High accuracy: Moderation (Audio) uses optimized deep learning algorithms to improve the voice moderation accuracy in complex environments.
    • Recognition of special sounds: Moderation (Audio) provides models for recognizing special sounds, such as copulation calls, moaning, and sensitive voiceprints.

Video Moderation

Application scenarios are as follows:

  • Video platforms/communities: Moderation (Video) accurately identifies non-compliant video content to help platforms and communities avoid risks.
    • Comprehensive detection: Moderation (Video) provides a multi-modal comprehensive review solution to parse the images, sounds, and text in videos.
    • Various types of video formats: Moderation (Video) supports video formats such as AVI, FLV, MP4, MPG, WMV, MOV, RMVB, and M3U8.
  • Video chat: Moderation (Video) accurately identifies and intercepts pornography, abuse, terrorism-related content, and advertisements in social and instant messaging scenarios.
    • Comprehensive detection: Moderation (Video) provides a multi-modal comprehensive review solution to parse the images, sounds, and text in videos.
    • Various types of video formats: Moderation (Video) supports video formats such as AVI, FLV, MP4, MPG, WMV, MOV, RMVB, and M3U8.
  • Online education: Moderation (Video) accurately identifies and intercepts non-compliant content in online teaching, interaction, and recorded courses to protect the physical and mental health of users, especially minors.
    • Comprehensive detection: Moderation (Video) provides a multi-modal comprehensive review solution to parse the images, sounds, and text in videos.
    • Various types of video formats: Moderation (Video) supports video formats such as AVI, FLV, MP4, MPG, WMV, MOV, RMVB, and M3U8.

Audio Stream Moderation

  • Live audio rooms

    Audio Stream Moderation is integrated into the Live audio platform to identify non-compliant content in live audio rooms in real time.

    Advantages:

    • Real-time: The content in live audio rooms can be monitored and analyzed in real time to ensure the order and security of the rooms.
    • Recognition of special sounds: Models for recognizing special sounds are available, such as asthma, moaning, and sensitive voiceprints.
  • Social voice messages

    Audio Stream Moderation reviews voice messages sent by users on the social voice message platform in real time, identifies voice messages that contain malicious content in a timely manner, and helps you take action based on the review result, for example, deleting messages or forbidding users to speak.

    Advantages:

    • High accuracy: All scenarios are covered, preventing misoperations or omissions and defending against risks in real time.
    • Recognition of special sounds: Models for recognizing special sounds are available, such as asthma, moaning, and sensitive voiceprints.
  • Online education

    Based on the education content and requirements, you can set appropriate review rules to help you identify audio streams that contain sensitive words or improper content, and detect and handle non-compliant content in a timely manner.

    Advantages:

    • High review efficiency: This service reduces manual review workloads, improves the accuracy of teaching content, and prevents incorrect or improper comments.
    • High accuracy: This service filters out inappropriate content and comments to ensure the security of teaching content.