CrazyEngineers : Voice Controlled Personal Assistant using Raspberry Pi

Made by: Akshay Nagpal (

Link to my project entry:

The objective of the project was to make a standalone personal assistant that can be interacted solely through the user’s voice. This project is a prototype for a variety of uses. It can help the user in doing simple tasks like checking mail inbox, time, weather conditions etc. as well as complex tasks like computing arithmetic problems, searching Wikipedia and face recognition for security at home. The user calls the system by speaking a keyword (“Cyrus”) through the microphone, following which the system signals the user (through a beep) to speak the task he wishes to accomplish. The user speaks the voice command and the system gives the output through the speakers attached to the system. The core concepts used in the project are speech-to-text conversion (for understanding user input) and text-to-speech conversion (for giving output to the user).

The project has features pertaining to information, entertainment and security. The user can check his Gmail for unread messages, ask the current time and calculate basic arithmetic problems using just voice commands. In addition to this, the system can search Wikipedia for any term that the user asks, and gives the results through voice only. The user can also entertain himself by telling the system to play music from his playlist in both normal and shuffled manner. For security, the system can click a picture through the webcam, and detect the face in the image. It further processes the face and tells whether the person is a friend or stranger. It does so by matching the face to a set of faces, already stored and labelled as friends in the system. OpenCV and normalized cross-correlation has been used for face detection and matching.

It can help the visually impaired to connect with the world by giving them access to Wikipedia, Calculator, Email and Music all through their voice. The voice controlled calculator can be used for visually impaired students in primary school. The prototype can also keep people secure as it can be used as a surveillance system which captures the face of the person standing at the door, and tells the owner whether the visitor is a known person or not.

– The project was written in Python programming language as it is the official language used for programming the hardware that I used, which was Raspberry Pi.


Facebook’s Caffe2 AI tools come to iPhone, Android, and also Raspberry Pi

Your smartphone may soon have the ability to recognize things in pics devoid of accessing the cloud

New intelligence can be added to mobile phones for example, the apple iphone, Android OS products, and low-power computer systems such as Raspberry Pi with Facebook’s new open-source Caffe2 deep-learning framework.

Caffe2 can be used to program artificial intelligence features into smartphones and tablets, allowing them to realize pictures, video, textual content, and speech and be more situationally aware.

It’s important to be aware that Caffe2 is not an Artificial intelligence program, but a tool allowing AI to be programmed into mobile phones. It takes just a couple lines of code to write learning models, which can then be bundled into applications.

The release of Caffe2 is significant. It indicates users will be in a position to get image recognition, natural language processing, and computer vision right on their smartphone. That work is commonly offloaded to remote servers in the cloud, with mobile phones then connecting to it.

Mobile gadgets are having more artificial intelligence abilities. More mobile phones are being bundled with Amazon’s Alexa and Google Assistant, while Apple’s Siri has been a staple in the iPhone for many years. Samsung’s Galaxy S8 mobile phones are due to get the Bixby voice assistant, which ought to make operating the handsets far simpler.

Caffe2 can function within the power constraints of mobile gadgets. It works with mobile hardware to hasten up AI apps and create neural networks.

Caffe2 uses the computing power of new mobile hardware to quicken deep-learning jobs. For instance, in mobile phones, Caffe2 will harness the computing power of Adreno GPUs and Hexagon DSPs on Qualcomm’s Snapdragon cellular chips.

Get more information at

Leave a Reply

Close Menu