UION-XERP Chinese
Sign inBook a demo
Menu
Order intakeAI imagesDocumentationPricingBook a demoSign inAccount & team

AI photo processing for personalized products

Prepare buyer photos.Make them fit your products.

Isolate people or pets, retouch naturally, restore photo quality or extend the image as needed. Configure the treatment in the product design scheme; XERP understands each photo, edits it and fills the result into its assigned position—without switching image tools for every order.

Built-in models · no connection setup

Select the treatment and get started, or connect compatible AI services and models of your own.

Discuss a sample photo →Explore the settings
Photo-specific treatmentProduct-scheme compositionHuman review
Concept illustration: the same golden retriever in a home photograph, isolated from the background, then printed on a dog-shaped fabric cushion
  1. Buyer photo
  2. AI subject extraction
  3. Product composition
AI prepares the photo; the design scheme composes the product artwork.Generated capability illustration, not a customer order or a measured AI result.

Start with the photo

Select the treatment, not every effect.

Photo processing and product layout are separate. Apply AI only to the chosen image material.

Portrait / half-body / precise cutout

Isolate the person or pet

Select the required subject area when placing a photo against another design background.

What to check

Check hair, clothing, subject edges and transparency.

Natural retouching

Keep the person natural

Refine appropriate portrait details without changing identity or trying to replace the face.

What to check

Check the expression and recognizable natural features.

Image restoration

Improve limited source quality

Improve detail where possible. Results depend on the input and cannot guarantee recovery of information absent from the original.

What to check

Check faces, texture and edges; do not assume invented detail is real.

Use expansion when the layout needs additional image area. Available combinations follow the actual configuration options.

Choose the photo treatment for your product

Set the photo placement and treatment once.

Choose where buyer photos go in the product design, then select the required isolation, retouching, restoration or expansion. New orders follow those settings. Built-in models are ready to select; custom connections are optional.

01 / Photo source

Place the right buyer photo in the design

Choose the buyer photo for each photo area. Fill the region with proportional cropping, or keep the size, angle and position from the buyer’s design. One photo can also be used in several positions.

Actual photo-source, display and condition controls in the Chinese interface
Source and placement also work without AI.
02 / Treatment

Choose the treatment for the intended result

Isolate a person or pet where the product needs the subject alone. Restore unclear detail, expand a photo that needs more image area, or retouch a portrait naturally. Available combinations follow the controls; isolated subjects can also use a configured outline.

Actual optional AI treatments and outline controls in the Chinese interface
Precise cutout is selected here. Subject-extraction modes are mutually exclusive and do not combine with expansion.
03 / Models

Use built-in models or connect your service

Select an available built-in model without entering an endpoint or key. If you have your own AI service, add the connection and choose models and fallback order. Keep product requirements configured and inspect the result with either service; confirm AI costs separately.

Optional custom connection with an example endpoint and empty key field
Chinese interface shown. Custom connections are optional, not a prerequisite.
Set photo treatments →Optional: connect your AI service →Repair artwork during review →

A photo-specific execution flow

Analyze → build an instruction → edit → refill.

The vision model reads this photo and the slot requirements. The editing model receives a tailored instruction, and the result returns to the selected position. Production dimensions and text layout remain controlled by the scheme.

  1. 01

    Read the photo

    Identify its content and the product’s actual needs.

  2. 02

    Prepare an instruction

    Combine photo-specific guidance with configured effects.

  3. 03

    Edit and refill

    Process the image and return the result to its slot.

  4. 04

    Review the composite

    Inspect identity, edges, crop, text and product layout.

AI results vary. If processing fails or is unsuitable, retain or restore the original and review manually. Confirm AI usage costs separately from artwork delivery charges.

Read pricing and cost boundaries →