[Digital Today reporter Daegeon Seok (석대건)] Musinsa said on 29 it has unveiled "Musinsa Lens," an AI image search service that lets users find products with a single photo. It described the service as expanding a text-led search experience into image-based browsing.

When a user taps the lens icon in the search bar and uploads a photo, the AI recognises products in the image. The results show identical and similar products available for purchase on the Musinsa store. The service supports major categories such as outerwear and shoes, and is available to members logged into the Musinsa app.

Users can find products based on visual information through Musinsa Lens, including fit, mood and colour tones that are difficult to translate into search terms. When users see a product they like on social media or offline, they can start browsing with only a photo, without needing exact search terms.

The service uses an AI model developed directly by Musinsa's tech organisation. The company said it trained its own detection model to find products in photos and another model to judge visual similarity between products. It can recognise up to 8 products in a single photo separately. The original photo and the searched products can be shared with a single link, allowing the recipient to see the same screen.

Musinsa plans to further advance its in-house image-recognition AI model to improve search accuracy and expand the services it is applied to. It will add search results filtering and sorting features in the fourth quarter and is pushing to introduce image search to 29CM as well.

A Musinsa official said the service follows the "AI Trend Curation" it introduced in June and is Musinsa's AI-based browsing function. The official said the company plans to broaden the browsing experience beyond AI first proposing products to include customers directly finding products through images.