General Usage
How does AI Studio determine which assets are eligible for annotation?
Only assets with a meaningful visual context can be annotated. This includes images, 3d models, and other assets with embedded or custom previews, Substance materials, and photometry assets (*.ies).
Assets without a usable visual context, such as Revit families or material libraries, require custom previews to be eligible. Learn more about custom previews in Connecter.
Can an asset without a visual context still be annotated?
Yes, assets without a visual context can still be annotated if you assign custom previews to them or generate previews for file formats like 3ds Max (.max), Cinema 4D (.c4d), and Revit families (*.rfa) using dedicated services.
Can I annotate assets stored in ZIP archives?
No, assets contained in ZIP archives aren’t supported for annotation.
Annotation Process
How long does the annotation process take for a large library?
The time depends on the number of assets and system resources. Annotating a large library can take several hours or even days, but metadata is available as it’s generated, so you can begin using it before the process finishes.
Can I stop and resume the annotation process?
Yes, you can stop the annotation process at any time. All metadata generated up to that point will be saved and immediately available in your library. When you restart the process, AI Studio will automatically skip previously annotated assets and continue with the remaining ones, ensuring no progress is lost.
How are assets handled if I restart an annotation process?
AI Studio automatically skips assets that have already been annotated in previous runs and processes only new or updated assets.
Can I use Connecter while the annotation process is running?
Yes, you can continue working in Connecter. However, running the annotation process on the same system may impact performance. Adjusting the process priority (affinity) can help balance system usage.
What happens if the annotation process is interrupted or stopped?
If the process is stopped, all generated metadata up to that point is saved and available in Connecter. When restarted, the process will skip already annotated assets and continue with the remaining ones.
Generated Metadata
Can I edit or delete AI-generated labels and descriptions?
Labels can be managed using the AI Studio Labels tool, where you can remove labels or prevent specific ones from being generated in the future. AI-generated descriptions can’t be edited or deleted.
How accurate are the generated metadata, and what factors affect accuracy?
Accuracy depends on the quality of the visual context. High-quality previews or images result in better annotation metadata, while low-resolution or missing previews may lead to incomplete or less accurate annotations.
While AI Studio is designed to deliver highly accurate annotations, it’s an AI-powered system, and there may be cases where the generated metadata is not entirely accurate or relevant.
Is the metadata generated by the Annotation Genie and Scheduler the same?
Yes, both tools generate identical metadata with the same quality, accuracy, and performance. The choice between the tools depends on your workflow preferences – use the Genie for manual runs and the Scheduler for automation.
What if my assets are stored in ZIP archives?
AI Studio cannot annotate assets stored in ZIP archives. Extract the files to a folder in your library, and then run the annotation process on the extracted files.
Custom Previews and Dedicated Instances
Do I need to enable the use of dedicated instances to annotate assets?
No, dedicated services are optional. However, they are useful for assets that lack custom previews or have low-quality embedded previews, as these may result in less accurate annotations.
If a custom preview is available, it’ll always be used, even if the related dedicated service is enabled. For assets with no visual context at all, enabling dedicated services is necessary to generate previews; otherwise, these assets can’t be annotated.
What happens if a required dedicated service is unavailable?
If a required service is unavailable, Connecter will rely on existing embedded previews or custom previews. If none are available, the affected assets can’t be annotated.
Can I use custom previews instead of embedded ones for annotation?
Yes, custom previews can be assigned to assets without embedded previews, and AI Studio will use these custom previews for annotation.
Cost and Balance
How is the cost of annotation calculated?
Annotation costs are based on the number of annotated assets and some other secondary factors. For example, annotating 10,000 assets may cost around 20.00 EUR.
Due to the nature of AI-powered annotation, providing a 100% accurate cost estimate in advance is not possible. However, you can view an estimated cost before starting the process and monitor the actual costs while the annotation is running. This allows you to stay informed and manage your prepaid balance effectively.
How is the annotation service billed?
The annotation service uses a separate prepaid balance system, independent of subscription cost for Connecter Suite.
You can add funds to your prepaid balance at any time, and the cost of each annotation process is deducted from this balance. Make sure your balance has enough funds before starting the process to avoid interruptions.
What happens if my prepaid balance runs out during the annotation process?
If your balance runs out, the annotation process will stop. Add funds to your prepaid balance and restart the process to annotate the remaining assets. Use the Cost Calculator to estimate expenses beforehand and avoid interruptions.
Can I limit spending on AI annotation?
Yes, you can manage spending by selecting specific root folders or asset types to annotate and monitor costs using the Cost Calculator before starting the process.
Tagging and Organization
How do I create tags or aliases from AI-generated labels?
Use the AI Studio Labels Tool to select one or more labels. You can create tags directly from labels or assign labels as aliases to existing tags for better organization.
Can I remove a label from all the assets to which it’s applied?
Yes, you can remove a label entirely from the system using the “Remove and Forget” option in the AI Studio Labels Tool. This deletes the label from all associated assets and prevents it from being generated again.
To remove a label from specific assets without deleting it system-wide:
- Single selection: Select an asset, open the “AI Studio Labels” section in the Asset Details Panel, and remove the label.
- Multi-selection: Select multiple assets, right-click, choose “AI Studio Labels” from the context menu, and remove the label from the group.
Language and Localization
Can I choose a language for the annotation metadata?
Yes, you can select the annotation metadata language during the annotation setup. Currently supported languages are English, Español, Deutsch, Français, Italiano, 日本語, or 简体ä¸ć–‡.
What happens to existing annotation metadata if I change the language in a new annotation run?
Existing annotation metadata remains unchanged. Only newly annotated assets will use the updated language.
Error Handling and Support
How can I monitor errors or failed annotations?
Use the Process Manager or the Annotation Scheduler to view error details. Address issues such as missing previews or unavailable dedicated services, then restart the process. Here are some of the most common errors:
- Error: Missing Previews -> Solution: Add a custom preview or enable dedicated services.
- Error: Service Unavailable -> Solution: Reconfigure the dedicated service or check the host application.
How is my data used to generate annotation metadata?
Connecter uses only the visual previews of your assets for AI annotation, not the assets themselves or their filenames.
These previews are sent securely to the AI service for analysis and are deleted immediately after use. They are never shared with third parties or used to train AI models. No personal or organizational information is accessed or shared. For more details, see our Terms of Use.