AI Features in Asset Manager

AI Features in the Asset Manager

The Asset Manager comes with a set of AI-powered tools that help you work with your media files more efficiently. You can automatically analyze images, generate brand-new images from text descriptions, edit existing images with natural language instructions, and create subtitles for audio and video files — all without leaving the Asset Manager.

These features are available when your organization has AI features enabled. You'll recognize AI-powered actions by the sparkle icon throughout the interface.

At a Glance

The Asset Manager includes five AI features:

  • Image Analysis — Automatically generates tags and descriptions (alt-texts) for your images
  • Image Generation — Creates new images from a text description
  • Image Editing — Modifies existing images based on your instructions
  • Subtitle Generation — Creates subtitles from spoken audio in videos
  • Audio Generation — Creates voice-overs directly in the asset manager

Image Analysis

Image Analysis uses AI to scan your images and automatically generate tags (keywords) and descriptions (alt text in multiple languages). This saves you from manually tagging every image and ensures your media library is well-organized and searchable.

Supported File Types

Image Analysis works with PNG, JPEG, WebP, and GIF files. SVGs, videos, and audio files are not supported for analysis.

Analyzing a Single Image

  1. Open any supported image in the file detail view.
  2. Click the "Analyse image with AI" button (sparkle icon).
  3. A confirmation dialog appears — click Confirm to start.
  4. A progress bar shows the analysis status. The process usually completes within a few seconds.
  5. Once done, the generated tags and description appear in the file's metadata section.

Analyzing Multiple Images at Once

You can also analyze images in bulk — either by selecting files or by analyzing an entire folder.

Selecting multiple files: Select two or more images in the Asset Manager. An "Analyse images with AI" button (with a sparkle icon) appears in the toolbar at the top. Click it to open a confirmation dialog showing the number of images to be analyzed, then confirm to start.

Analyzing a whole folder: When you're viewing a folder that contains unanalyzed images (and you haven't selected any files), the same "Analyse images with AI" button appears in the toolbar automatically. Clicking it will queue all supported images in the folder — including images in subfolders — for analysis. A confirmation dialog shows the total number of images before you proceed.

On smaller screens, the button appears as a compact sparkle icon instead of a labeled button.

What You Get After Analysis

After analysis, each image receives:

  • Tags — A set of descriptive keywords that describe the image content
  • Descriptions — Auto-generated alt text, available in multiple languages
  • An "AI-analyzed" badge visible in the file grid, showing the number of tags

The AI-generated descriptions are labeled with an "AI" source tag so you can always tell them apart from manually written descriptions.

Why Analysis Matters

Image Analysis is also the first step before you can use AI Image Editing. An image must be analyzed before the editing feature becomes available. If you try to open the image editor on an unanalyzed image, you'll see a tooltip: "This image needs to be analysed first before AI editing is available."

AI Image Generation

The Image Generator lets you create entirely new images from a text description. Describe what you want, choose a format, and the AI produces an image you can save directly to your Asset Manager.

Opening the Image Generator

  1. In the folder tree on the left side of the Asset Manager, click the "New" button at the top.
  2. A dropdown menu appears with options for creating new content. Select "AI Image" (shown with an image icon).
  3. The Image Generator dialog opens with a split view: your settings on the left and a live preview on the right.

Creating an Image

  1. Write a description — In the "Image description" text area, describe the image you want to generate. Be as specific as possible for best results. The description must be at least 4 characters long.
  2. Choose an image name — Enter a file name for your new image (e.g., "galaxy_edition_hero"). The name must be at least 4 characters and can only contain letters, numbers, spaces, dots, dashes, and underscores. Don't include a file extension — it's added automatically.
  3. Select a format — Choose one of three aspect ratios:
    • Landscape (1536 × 1024, 3:2 ratio)
    • Portrait (1024 × 1536, 2:3 ratio)
    • Square (1024 × 1024, 1:1 ratio)
  5. Click "Create image" to start generation. A loading indicator appears in the preview area while the image is being created.
  6. Once the image appears in the preview, you can review it. The dimensions are shown as a badge on the image.

Saving or Discarding

  • Click "Upload image" to save the generated image to the current folder in your Asset Manager.
  • Click "Close" to discard the image and close the dialog.

If you're not happy with the result, simply adjust your description and click "Create image" again to generate a new version.

AI Image Editing

AI Image Editing lets you modify existing images using natural language instructions. Instead of opening an external editor, you can describe the change you want — like "remove the background" or "make the sky more dramatic" — and the AI applies it.

Requirements

Before you can edit an image with AI, it must first be analyzed (see Image Analysis above). The edit button only becomes active after analysis is complete.

Opening the Editor

  1. Open a supported image in the file detail view.
  2. Click "Edit image with AI" (sparkle icon).
  3. The editor opens as a full-screen dialog with three columns.

The Editor Layout

The editing interface is divided into three areas:

  • Left pane — Source image: Shows the original image or the version you're editing from. The label indicates whether it's the original or a previous edit.
  • Center pane — Edit area: This is where you describe your edit and see the result.
    • Type your edit instruction in the text field (e.g., "Make the background blue" or "Add a subtle shadow")
    • Click the send button to start generating
    • The AI result appears below once ready
  • Right sidebar — Versions: Shows all your edit versions as a scrollable list of thumbnails.

Editing Step by Step

  1. The original image is shown on the left. Type your desired change in the prompt field in the center.
  2. Click the send button (paper plane icon). The AI processes your request — this may take a few moments.
  3. The result appears in the center pane. Review whether it matches what you wanted.
  4. To make further edits, you can:
    • Type a new prompt to edit the result again
    • Click "Edit from this version" on any previous version in the sidebar to use it as the starting point

Managing Versions

Every edit you make is saved as a version in the right sidebar. You can:

  • Browse versions — Click any thumbnail to view it
  • Edit from a specific version — Click "Edit from this version" to use it as the source for your next edit
  • Delete versions — Click the trash icon to remove versions you no longer need

The original image is always available at the top of the version list and cannot be deleted.

Saving Your Edits

When you're happy with a result, you have two options:

  • Save as new file — Creates a new image file in the Asset Manager, keeping the original unchanged
  • Overwrite original — Replaces the original file with your edited version

Both save options are only available after you've generated at least one edit.

AI Subtitle Generation

AI Subtitle Generation automatically creates subtitles from the spoken audio in your video or audio files. The AI detects the source language, transcribes the speech, and generates a subtitle track in your chosen target language.

Opening the Subtitle Manager

When viewing a video or audio file in the file detail view, you'll see a Subtitles section in the sidebar. This shows all existing subtitle tracks and lets you add new ones.

Generating Subtitles with AI

  1. Click "Add subtitle" to open the subtitle dialog.
  2. Select the "Generate with AI" tab (it may already be selected by default).
  3. The source language is shown as "Auto-detected from audio" — the AI figures this out automatically, so you don't need to set it.
  4. Choose a target language from the searchable dropdown. Type to search through available languages. Languages that already have a subtitle track are excluded from the list.
  5. Click "Generate subtitles" to start the process.

While Subtitles Are Being Generated

Subtitle generation typically takes about one minute. You don't need to wait — the dialog tells you: "Generation usually takes about a minute. You can keep working — the track appears here when it's ready."

While generating, the subtitle appears in the list with a spinner and a "Generating..." label. Once complete, it becomes a fully usable subtitle track.

Managing Subtitles

Each subtitle in the list shows its language and an origin badge:

  • AI (sparkle icon) — Generated by AI
  • Uploaded (cloud icon) — Manually uploaded

For each subtitle, you can:

  • Edit — Change the subtitle's name. This name is used as the label that appears in the subtitle selector in the Player, so choose something recognizable for your learners (e.g., "English" or "English (auto-generated)").
  • Download — Save the subtitle file to your computer
  • Delete — Remove the subtitle permanently (with confirmation)

If you have more than 6 subtitle tracks, the list is collapsed by default. Click "Show all" to expand it, or "Show fewer" to collapse again.

Uploading Subtitles Manually

If you prefer to use your own subtitle file, switch to the "Upload file" tab in the Add Subtitle dialog to upload an existing subtitle file instead of generating one with AI.

Audio Generation

View the article on audio generation with AI here.

Tips and Best Practices

  • Analyze images first — Running Image Analysis on your library gives you searchable tags and enables AI editing. Consider analyzing all images in a folder at once using the toolbar button.
  • Be specific with prompts — Whether generating or editing images, detailed descriptions produce better results. Instead of "a nice landscape," try "a panoramic mountain landscape at sunset with warm orange and purple tones."
  • Use versions wisely — When editing images, you can build on any previous version. Try different approaches and pick the best result before saving.
  • Generate subtitles early — Since subtitle generation runs in the background, start it early and continue working on other things while it processes.
  • Choose the right format — For image generation, pick the aspect ratio that matches where you'll use the image. Landscape works well for headers and banners, portrait for mobile-first content, and square for thumbnails and social media.


Was this article helpful?

English
Powered by Product Fruits