First we express our deepest thanks and gratitude to the almighty God for his divine blessing enabling us to complete the last year project/internship successfully. Deep knowledge and great interest from our supervisor in "NLP & IOT" to carry out this project. Fahad Faisal, Shams Sunny and Head, Department of CSE, for his kind help in completing our project and also to other faculty members and staff of CSE Department of Daffodil International University.
We would like to thank all our coursemates at Daffodil International University, who participated in this discussion while completing the coursework. Finally, we must acknowledge with due respect the constant support and patience of our parents. Now various types of interactions are being developed that make using our devices easier than ever.
It is now very popular in the current era because all the tech giants are focusing on speech recognition with their. But unfortunately all these jobs are done in English or other popular languages.
Motivation 1
The smart home lets you control a number of connected devices in your home from a smartphone [1].
Objectives 2
Expected Outcome 2
Report Layout 2-3
In the back-end design, we discuss how CMU Sphinx works, the app configuration, and the microcontroller configuration. In this chapter we discuss the implementation of the database, Implementation of Front-end design, Interaction and testing. This chapter is about the conclusion and the possibilities for further development of this project.
We reviewed several works on Bangladeshi speech recognition and some works on home automation. So we wanted to take those works, improve them and use them in a real-life scenario.
Related Works 4
Comparative Studies 4-5
After completing the project, a new horizon will open for speech recognition with the Bengali language. People can contribute to the speech recognition to improve the accuracy and detection of the Bengali command. Another extent, this smart home help will be cheaper than other smart assistance and it will compete with the assistance of the others.
Challenges 5
In this process we can see that at first, the user has to record the voice, then the voice goes to CMUSphinx. If the saving mode is on the command is stored in the database and if bluetooth is enabled it will try to find a match from the database. If a match is found, it will send the command (ON/OFF) and the equivalent pin no to the Arduino.
The figure shows from the beginning, where the user gave the command, to the stage where the command was executed. The output is further processed and a command is issued to the microcontroller to excite the command. An example of use is the methodology used in system analysis to identify, clarify and organize system requirements [13].
By using the use case model, we are able to better understand possible errors, errors in software processes. If the user turns Bluetooth on/off, the app will detect and act according to the state. If Bluetooth is turned off, it will prompt the user to turn on Bluetooth to send data.
Description: After processing the output from CMUSphnix, the system will check if save mood is ON/OFF. Description: After processing the output from CMUSphnix, the system will check whether Bluetooth is on or not. If a match is found in the database, the PIN will be sent to the microcontroller via Bluetooth.
The different steps such as configuring the transcription, file IDs of audio sources to the final generation are shown here.
Design Requirements 17
DESIGN SPECIFICATION 18-31
Implementation Requirements 30-31
Also, this system uses an Arduino Uno to execute commands sent by the app via Bluetooth. CMU Sphinx is a framework created by Carnegie Mellon University to provide the necessary tools for building an offline speech recognizer. Pocket Sphinx is a tool that is the simplest version of the decoder to provide speech-to-text output with minimal resource usage.
After finishing the user's speech, we will check if it matched the database. If it matched with the database, we check the trigger command and then send the pin and trigger command to the Arduino.
Implementation of Front-end Design 32-33
If the user wants to save any command, a popup window will open and ask for the pin number. Users can also check the list of saved commands by pressing the list button at the top left of the lunch activity. We have tried to make the application interactive with the user so that the user has to put less effort in a single operation.
So, when the user records the voice, the voice command is sent to the Arduino and the Arduino will execute the command according to the user's request. Users can also save a command to determine the pin number of household appliances to which PIN they are connected.
Testing Implementation 34-35
CONCLUSION AND FUTURE SCOPE 36
Discussion and Conclusion 36
Control your home appliances with just one hand using your smartphone and speak a Bengali voice command. In this home assistant, we use the CMUSphinx voice recognition framework and develop a mobile application to recognize Bengali voice commands.
Scope for Further Developments 36
We decided to create a home assistant that can effortlessly make home automation easy for everyone. Sinha, “Home Automation via Bluetooth Using Android Application”, ABHIYANTRIKI: ABHIYANTRIKI: An International Journal of Engineering &. Available: https://www.uxmatters.com/mt/archives/2012/01/defining-an-interaction-model-the-cornerstone-of-application-design.php.
