Pinterest unveiled a collection of fresh visual search capabilities grounded in artificial intelligence — consumers will have the ability to pinpoint an object of interest in an image and explore similar ones. Initially, these instruments will be accessible for content concerning women’s fashion in the USA, Canada, and the UK.
Pinterest employs visual language models (VLM), a variation of generative AI, to produce descriptions of styles and the overall vibe of pins. The new instrument dissects and deciphers images, to facilitate the hunt for specific details in outfits.
When browsing a pin, Pinterest’s AI proposes terms that may be employed to characterize diverse portions of the image, for instance, “overall aesthetic,” “color scheme,” “cut,” or “product category.” A fresh animated glow highlights objects with which one can interact. Pinterest appends a new filtering panel, which will assist consumers in narrowing search outcomes. For example, to a favored object, one can apply a “style” or “occasion” filter to discover an option that will suit any circumstance.
The fresh visual search capabilities are primarily intended for content concerning women’s fashion and are accessible in the USA, Canada, and the UK. Pinterest intends to launch this capability in other categories and countries over time.
