Multimodal application concept
Multimodal application
Application on Samsung TV:
A project that allows users to interact with the TV using AI technologies for voice and vision, focusing on shopping and recognizing celebrities. Products were recognized based on the scene on TV, and online searches were conducted to recommend sales sites to the user. In addition to recognizing the celebrity, a brief biography and their latest movies/series were also displayed.
Challenge
The primary challenge lies in implementing this tool within a live-streaming application for movies, videos, and similar content. The interface must be exceptionally clean and intuitive, as users have limited screen time and cannot pause the content. It's essential that the tool does not obstruct visibility, allowing users to understand and interact seamlessly while maintaining an optimal viewing experience.
Mobile Application:
A project that allows users to continue their shopping experience (viewing products they saved on TV more closely and visiting links to shopping sites). A large language model (LLM) was also used to redesign products with prompts (changing color, pattern, etc.). The redesigned product could also be searched online using technologies like Google Lens.