Speech AI

AI Audio Description

Add narrated descriptions of what happens on screen to your videos, so blind and low-vision viewers can follow every scene. Review the script, choose a voice and language, and export the result.

At a glance
  • Automatic description script
  • Fits the pauses in dialogue
  • Edit before voicing
  • Natural, multilingual voices

What is AI audio description?

Audio description is an extra narration track that describes the key visual elements of a video, such as actions, settings, people and on-screen text, for people who are blind or have low vision. AI audio description automates the work: the visual content is analysed, a description script is written into the gaps in the dialogue, voiced with a synthetic voice and mixed with the original soundtrack.

Input
Video files (for example MP4, MOV or MKV)
Output
Mixed audio track and description script
Review
Edit the script before it is voiced
Languages
Narration in many languages and voices
Content
Films, series, e-learning, marketing and corporate video
Access
Klassifier web app
The challenge

Much of a video is told visually

Actions, settings, facial expressions and on-screen text carry meaning that dialogue alone doesn't convey. Writing, recording and mixing audio description by hand takes time and budget, so many videos are published without it.

Our solution

Description written, voiced and mixed for you

Klassifier analyses your video, writes short descriptions of the important visual information and places them in the natural pauses between dialogue. You review and edit the script, then it is voiced with a natural AI voice and mixed with the programme audio.

Features

Automatic description script

The visual content is analysed and described in short, factual sentences: who is on screen, what they do, where the scene takes place and any important text.

Fits the pauses in dialogue

Descriptions are timed to the gaps between spoken lines, so the narration doesn't talk over the dialogue.

Edit before voicing

Choose script-only output, review and correct the description in an editor, then render the voiced version once you are happy with it.

Natural, multilingual voices

Pick the narration language and a natural AI voice from the Klassifier voice library.

Context makes it better

Add an optional synopsis and a list of the main characters so the narration names people correctly and understands the story.

Ready-to-use exports

Download the audio track with the description mixed into the programme sound, and the description script for your own review or archive.

How it works

From video to described video in four steps

  1. Upload your video

    Add the video in the Klassifier web app and choose the narration language. Optionally add a synopsis and the cast.

  2. Get the script

    Klassifier writes descriptions of the visual content and places them in the pauses between dialogue.

  3. Review and edit

    Read through the script, adjust wording or timing, and choose the voice.

  4. Voice, mix and export

    The script is voiced and mixed with the programme audio. Download the mixed track and the script.

Use cases

Films and series

Make drama, documentaries and series accessible to blind and low-vision audiences on your platform or channel.

E-learning and training

Describe slides, demonstrations and on-screen text so every learner can follow the course.

Marketing and social video

Reach a wider audience with campaign, product and brand videos that work without seeing the picture.

Corporate and public sector

Add description to internal communications, public information videos and recorded events.

Benefits

Helps you meet accessibility requirements

European Accessibility Act

The EU European Accessibility Act sets accessibility requirements for a range of products and services, including some audiovisual media services. Audio description helps you make video content accessible to blind and low-vision users.

WCAG 2.x

The Web Content Accessibility Guidelines include success criteria for audio description of prerecorded video. A description track produced with Klassifier helps you work towards them.

FCC and CVAA in the US

In the United States, the FCC's audio description rules under the CVAA apply to certain broadcasters and video programming distributors. Klassifier helps you produce description tracks efficiently.

Your responsibility, our tooling

Requirements depend on your sector, country and content. Klassifier is a production tool and does not certify compliance; review the output and check your obligations with your accessibility or legal advisor.

Audio description: frequently asked questions

What is the difference between audio description and subtitles?
Subtitles and captions turn the audio into text for people who can't hear it. Audio description does the opposite: it turns important visual information into narration for people who can't see the picture.
Can I edit the description before it is voiced?
Yes. You can stop after the script, review and correct it, and only then render the voiced and mixed version.
Which languages are supported?
Narration can be produced in many languages with voices from the Klassifier voice library. The languages available for your project are listed in the web app.
What do I get at the end?
An audio track with the description mixed into the programme sound, and the description script.
Does using Klassifier make my content compliant?
Klassifier helps you produce audio description, which is part of many accessibility requirements, but it does not guarantee or certify compliance. Obligations differ by country, sector and content type, so check them with your advisor.
Should a person review the result?
We recommend it. AI-generated descriptions are a strong first draft; a human review makes sure names, tone and important details are right before you publish.

See it with your own content

Start a free trial in the Klassifier web app, or book a demo and we'll walk you through it with your own data.