Hello. I saw some videos about esp-eye which is programmed via esp-idf and I realized that to record a face, create a faceID, it is necessary to be connected to a PC to generate the video stream on the screen.
But after recording the faceID, to carry out recognition, is it also necessary to be connected to a PC or does it work autonomously?
I changed the title so that this space can be expanded by anyone who wants to contribute with useful information and especially codes, or links to codes about ESP-EYE.
To answer your first question, after recording a face ID, the ESP-EYE should be able to perform face recognition autonomously, without the need for a PC connection. However, it's important to note that the ESP-EYE is a development board designed for prototyping and experimentation, and the level of performance and reliability may depend on various factors such as the quality of the camera, lighting conditions, and processing power of the ESP32 chip.
As for your second question, the number of faces that can be recorded may depend on the available memory of the ESP32 chip and the size of each face ID. However, according to the ESP-EYE datasheet, it has a built-in hardware accelerator for face detection and recognition, and it can support up to 2000 face IDs. It's also possible to optimize the storage usage by using techniques such as face embedding, where a compact representation of the face is used instead of the raw image data.
I hope this helps! Let me know if you have any further questions.