Introduction: To Realize Desktop AI With Raspberry Pi 3b
To realize full-desktop control，A high precision gesture recognition and speech synthesis system based on raspberry pi is established.
With the passion for smart home, I like to upgrade and transform the things around me.As a junior maker, I believe that will change our lives step by step.
n this project, 24 gestures can be recognized with high precision by raspberry pi, and different desktop instructions can be made for gestures, and voice synthesis broadcast can be carried out to achieve full control of the desktop.
Next, we will be divided into 3 parts. First, through gesture control, the web crawler returns information.This time we are operating the weather station. Through five instructions, we can understand the temperature in the next five days and carry out voice synthesis and broadcast. Second, we can control other operating systems through gestures, and what we do this time is to power on the computers in the LAN. Third, we are familiar with the interior of the raspberry pi, by getting the current time of the raspberry pi, and then calling the speech synthesis api for speech synthesis and output.
Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.
Step 1: Assemble Hardware and Build IDE
The assembly of the hardware is very simple, so please focusing on the installation of drivers and libraries.
Step 2: Call Gesture Recognition Library Api
By comparing the AI recognition ability and openness of different manufacturers in the market, I chose Baidu's recognition library. So what you need to do in this step is to take a photo through the camera,
and then call Baidu's api interface (including three id and key) to use its sdk for image recognition, and a value will be returned after recognition. We use different commands by generating this value.
Step 3: Get Local Time
We need to network the raspberry pi first, and then get the local time directly.
Step 4: Step4：Wake-on-LAN
Wake-on-LAN is abbreviated to WOL or WoL. WOL is not only a technology, but also the standard of this technology. Its effect is to make the computer that has entered the dormant state or shutdown state issue orders toit through the other end of the local area network (mostly Ethernet), so that it can wake up from the dormant state, return to the operating state, or change from the shutdown state to the boot state. In addition, the technologies related to WOL also include remote control mechanisms such as remote order shutdown and remote order restart.
First of all, we need to set the computers that need to be awakened can be waked up driven by the network card. Then we connect the raspberry pie and the computer together in the local area network, and through the gesture transfer instruction, the transmission packet is assigned to the computer through the mac and ip address, thus enabling it to start.
Step 5: Catch Website Information
Using python to crawlers the weather website and carry the information out to the website to output the local weather conditions.
Step 6: Call Baidu's Speech Synthesis Api
In this step, we still call Baidu's speech synthesis api to increase the accuracy of audio output, which reduces a lot of time cost for us.