Image Recognition
Image recognition (also known as image analysis or vision AI) analyses digital images using artificial intelligence to recognise objects, text or faces.
Information on Data Processing
For instance, a screenshot is captured or a photo is taken and sent to the respective provider for analysis. Accompanying metadata of the images (such as capture date, device type, or location, if available) may also be used for contextualization. The provider may use the submitted data (images and prompts) to improve and train their models.
Use Cases
With many providers, image recognition is already integrated into their large language models, allowing the feature to be accessed directly through the chatbot.
Daily Life Assistance
Using an app, users can take a photo to have a menu translated and read aloud, or retrieve historical information about a building. These apps can also assist blind and visually impaired individuals with tasks such as reading transit timetables.
Automatic Scene Descriptions
The first frame of each scene is analyzed; in the future, summaries for individual scenes in the video may also be provided. The result is made available as a separate document and can be displayed in the editor alongside the manuscript being created.
Interactive Chatbot
It is possible to “chat” with an LLM directly within the editor. This allows users to ask questions about the current video frame, for example, to read name tags or on-screen subtitles.
Providers Integrated into Frazier
- Google Gemini: Website, Privacy Policy, Terms of Use.
- OpenAI Website, Privacy Policy, Terms of Use.
Last Update: 2026/07/08