Users often take screenshots from shopping websites/apps for later consideration for purchase. However, accessing the product page at a later time requires the user to manually enter the website/product name. This disclosure describes techniques that help users shop for products based on a screenshot of an online page containing information about the product. With user permission, it is determined whether an image on the device is a screenshot. If an image is determined to be a screenshot, a suitably trained machine learning model is employed to determine whether the screenshot contains a purchasable product and OCR is applied to detect whether the screenshot contains text that includes the product price. If a screenshot is identified as containing a purchasable product with a corresponding price, when the user subsequently accesses the screenshot, the shopping action is surfaced using a suitable user interface (UI) mechanism such as a tooltip or button. Interacting with the tooltip or button takes the user directly to the website or app where the product can be purchased.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.