Visual Attributes

Identify product attributes from images

Grounded in a standardized fashion taxonomy, visually identify and extract product attributes from any image in the catalog.

Retail customers today have high expectations for product information. Descriptive product attributes impact the consumer’s ability to effectively search (or browse) and find products online; narrowing down by color, fit, fabric, and so on. Managing an e-commerce product catalog is especially difficult for merchants selling thousands of SKUs from hundreds of suppliers across multiple sales channels: a lot of time and money is spent merchandising or enhancing product attributes. Merchants dread their attribution management duties, as it has become a mundane task separating them from their core responsibilities. This annoyance and industry challenge is further exacerbated by the growing adoption of natural language and voice interfaces, and chat bots, which are changing the digital search paradigm.

API documentation


Use cases

  1. Auto-attribute Generate product attributes from catalog images.
  2. Enriched attributes Gather additional attributes from product images and descriptions.
  3. Post-filtering of image search results