How to Generate More Views and Poses from a Try-On Image

Once you have one image that’s finalized and ready to go, you can use the Create more poses and views feature to create more variation from the final images you’ve selected.

Step 1: Select an approved image

Click into the review page from a task. Select the image that’s finalized with editing and ready to use. This way it’ll reduce the level of editing needed to the minimal.

Step 2: Click the change pose button

Click the Create more poses and views button from the upper right corner.

Step 3: Select Pose

You’ll see a pop up where you can multi-select the poses or customize the poses you want. Once you are done with selecting, click the generate x poses button to generate the poses.

Using this method to generate images, the output results’ resolution might be smaller than the original at 1.2K height. You can use the upscale feature in editing to update the resolution of the images.

Step 4: Review Results

Once submitted the selected poses, the results will be generated and stored in a separate task. You can click into each of the task to review the results.

If you are not satisfied with the results, you can click the regenerate button and regenerate the try on for more accurate results. Or use the editing feature to edit the image.

Pose Categories

The system offers 4 categories of model poses: Women, Men, Kid and Baby.

How to Customize Poses

You can also customize your own poses.

On the change pose pop up, toggle over to the Custom Poses on the top.

In the custom pose module, upload a reference pose image, select the pose crop type and click the Extract Pose button.

The system will automatically generate a sketch image based on the referenced pose.

This prompt will only extract the view, pose and crop of the pose and ignore any other details in the image. You can re-extract pose if the result is not satisfying. Each extraction will cost 2 credits.

Once saved, the custom pose will appear on the Custom Pose section.

Select the pose and you’ll be able to generate an image with the similar pose to it with the same model, clothing and background as the original try-on image.