The developer updated the Raspberry Pi 5-based robot to handle image processing and voice commands more effectively. The system now includes a preview feature for photos, allowing the user to verify the image quality before the robot analyzes it. Users can capture up to four images for a single query, with the ability to replace individual shots.
The robot now supports printing via a network-connected Epson L366 printer using the CUPS system. Commands like 'print' trigger a preview mode, allowing users to select image size and color before confirming the print. The interface also displays educational content, such as math solutions, in a step-by-step format on the screen.
To improve voice interaction, the developer optimized the audio settings to prevent the robot from speaking over the user. The system now prioritizes listening before initiating greetings. For language support, the project integrates Whisper tiny for English queries alongside the existing Vosk model for Russian, though automatic switching is not yet implemented.
The robot now features a local SQLite database on the Raspberry Pi to store user profiles and conversation history. This allows the system to recognize specific individuals and maintain context during interactions. Additionally, an on-screen keyboard was added to the settings menu to facilitate Wi-Fi configuration and profile management without external peripherals.
