FAQ

Q: Can I change the scene type through prompts, such as converting a Plane AR scene into an Image AR scene?

A: No. AI Enhancement works based on an existing AR scene, so prompts cannot change the scene type.Before using AI Enhancement, you need to create the appropriate AR scene type first. AI will then help you add interactions, animations, and other enhancements based on that scene.

Q: What should I do if I am not satisfied with the generated result?

A: AI-generated results may vary. If the result does not meet your expectations, you can continue refining it in the same conversation by adding more specific instructions.For example, if the generated object appears too large, you can add a prompt like:“Make the object smaller and reduce its size to one-third of the current size.”AI will then adjust the generated result based on your feedback.

Q: Can I create an AR scene from scratch using AI Enhancement?

A: No. AI Enhancement works based on an existing AR scene and is designed to help you add interactions, animations, and visual effects.Before using AI Enhancement, you need to create a scene first. You can place your assets in the scene and then use AI Enhancement to generate the desired interactions and animations.

Q: Can all AR scenes use AI Enhancement?

A: Currently, AI Enhancement is not supported for Plane AR, World AR, and Roaming AR scenes.Other supported AR scene types can use AI Enhancement to create interactive content and animations.

Q: How can I preview AR experiences created with AI Enhancement?

A: Currently, AI Enhancement experiences cannot be previewed through mini programs.You can use WeChat Scan or your mobile browser to preview the generated AR experience.

Q: Why didn’t the expected effect generate even though I described it in the Interactions prompt?

A: First, check whether the asset has been correctly referenced. When entering prompts in the prompt area, we recommend clicking the asset directly from the object list to reference it.After successful referencing, the prompt area will display the asset in the format “@Asset Name”. This ensures that AR Agent can correctly identify the target asset and generate the desired effect.

Q: Why does the generated AR experience not play the animation as expected after setting up asset animations with AR Agent?

A: This usually happens because the AI does not accurately recognize the interaction prompt you provided, or because the asset animation reference format is incorrect. Please check the following:

• Confirm that the animation reference format is correct After selecting an asset, use the format @Asset Name(Animation Name) in your prompt to reference the animation.
• Make sure the animation name matches exactly The animation name entered inside the parentheses must be exactly the same as the animation name in the editor scene, including capitalization, spaces, and special characters. Using the correct animation reference format helps AR Agent better understand your requirements and generate the expected interaction behavior.

Q: How can I improve the accuracy of AR Agent-generated content during the creation process?

A: To achieve higher-quality generation results, it is recommended to review the implementation logic provided by AR Agent after generation and optimize your prompt based on its feedback. You should mainly check:

• Whether the interaction logic matches your expectations
• Whether the asset names and animation names are correctly matched
• Whether the prompt description is clear and complete

If the generated result differs from your expectations, you can improve the accuracy by refining the prompt, adding more details, or correcting asset references. This helps AR Agent better understand your creative intent.

Q: When creating an AR scene with multiple complex interactions, how can I improve the production efficiency?

A: For complex interactive scenes, it is recommended to use a step-by-step generation approach.You can first describe and generate the first interaction feature. After confirming that the result meets your expectations, continue adding the next interaction. Gradually build the entire scene by completing each functional module. By breaking down complex requirements into smaller steps, AR Agent can better understand the logic of each interaction. This approach also makes it easier to adjust and optimize the experience, allowing you to create more complex and stable AR experiences.