AI Try-On
发布于：2025-12-05
Purpose:
Supports generating realistic human try-on images by using AI to place the completed 3D garment model onto the model shown in a reference photo.
Operation Path:
Click the AI Function Integration button at the top right of the software
▶ AI Try-On
Operation Instructions:
Reference Image：
You can perform AI virtual try-on for the current garment based on the provided reference image.The reference image can be selected from the software’s built-in model library, or you can upload any model photo from your local files for AI try-on.
When executing the AI try-on, the AI will dress the garment you created onto the model in the reference image according to the structure and design of the current garment.
AI Cloth：
Operation Instructions:
By adjusting the following parameters, you can make the garment details appear more realistic — for example, increasing the realism of wrinkle intensity in the generated clothing, controlling how many design elements are preserved, and configuring the material settings of the output garment.
Function Overview：
AI Coth |
| Depending on the selected strength level, the realism of the AI try-on results will vary. A higher strength value produces more realistic try-on effects, making the garment presentation clearer and more visually convincing. |
|---|---|---|
| Controls the level of garment details retained in the AI try-on output. The higher the preservation level, the more detailed the retained garment elements. Low: Retains zippers / buttons / patterns Medium: Retains topstitching / zippers / buttons / patterns High: Retains topstitching / textures / zippers / buttons / patterns | |
| Based on the differences in garment category settings, the AI try-on system will automatically adjust the clothing in the exported images according to the selected category. This enhancement strengthens the expected texture and wrinkle behavior that corresponds to each specific material category. For example, if the Fleece category is selected, the exported try-on image will be adjusted to reflect fleece-specific characteristics, such as fabric texture and material softness, making the AI-generated clothing appear more realistic and true to the expected look and feel of fleece garments. |
Image：
Function Overview:
Image | Model Type
| Model TypeThere are four model types: HQ1.0： Includes garment processing and an automatic hand-fixing option. The generated image contains both background and model. HQ2.0： Includes garment processing and an automatic hand-fixing option. The generated image contains only the model and clothing, without a background. FS1.0： Includes garment processing but does not include automatic hand fixing. The generated image contains both background and model. FS2.0： Includes garment processing but does not include automatic hand fixing. The generated image contains only the model and clothing, without a background. FS modes focus more on preserving the original pose of the person in the reference image for more accurate garment try-on. HQ modes focus on generating a more natural try-on effect on top of the reference image. |
| ||
| ||
Cloting Type
| Different clothing types will generate corresponding garment-wearing effects: Suits: Generates a complete set of clothing directly on the model. Tops: Generates only the top garment on the model while keeping the bottom part from the reference image unchanged. Bottoms: Generates only the bottom garment on the model while keeping the top part from the reference image unchanged. Note: When generating one-piece garments such as dresses, both the Set and Top options will produce a full-body try-on result. The Bottom option will apply only the lower part of the garment model to the try-on image, while the upper body retains the appearance from the reference image. | |
| ||
Hand Posture Refinement
| When enabled, the system automatically adjusts and refines the model’s hand appearance based on the state of the clothing. |
