​
Overview

Unbxd Visual Search allows shoppers to find products by uploading an image instead of typing a search query. It uses advanced AI and computer vision to analyze the image, extract relevant attributes (like color, category, texture, etc.), and surface visually similar products from your catalog in real-time.

​
Prerequisites

  1. You should have a valid UNBXD account with appropriate permissions to access the Self-Serve Console.

  2. Visual Search must be enabled on your UNBXD panel.

    Note:

    1. Visual Search is a PAID feature and is not included by default in the standard Unbxd Panel.
    2. Merchandisers do not have access to enable this feature, contact Unbxd Business Team to enable this feature for your panel.

  1. Navigate to the Search Preview of your console.

  2. In the Search Preview, click on the camera icon to open the Visual Search panel.

  3. You can choose your input in two ways for a search:

    • Upload an Image File from your device (Supports standard image files like JPEG, PNG, and so on)
    • Paste an Image URL to see similar items

  4. Once the image is processed the image is displayed on the left as the anchor image.

    It will include interactive markers (dots) on different product areas detected in the image.

  5. Click on any marker (dot) to view products similar to that item. For example: Click on a marker over shoes to see similar shoes from the catalog.

Markers help pinpoint specific items in a multi-product image. For example, If in an image, a model wearing a dress and shoes, you would see markers over both.

  • Click the dress marker to see similar dresses.
  • Click the shoe marker to see similar shoes.
  • Click on “Delete” icon to delete the current image and re-upload or provide a new image URL.
  • Each time you upload or paste an image, a new anchor image and markers will appear.

​
Conclusion

Unbxd Visual Search revolutionizes the way shoppers interact with your site by enabling search through images rather than text. By leveraging AI and computer vision, it helps users quickly discover products that visually match their preferences without the need to know specific product names or descriptions.

This feature is especially useful for fashion, home décor, and other visually-driven categories, where appearance plays a crucial role in purchase decisions. With support for both image uploads and URLs, Visual Search enhances product discoverability, reduces zero-result searches, and delivers a faster, more intuitive shopping experience.