Talha Mujahid

Academic

Career

Verbatim text

Automatic Speech Recognition System to Generate CV 
 
SpeakCV, is an automatic CV generating system for native Urdu speakers that uses speaker’s voice 
information as input and generates a professional CV using Neural nets and Natural Language Processing Techniques. 
The user will answer a few questions in Urdu and user’s speech data will be recorded in phone. The 
data is then pre-processed and sent to a speaker independent ASR for speech recognition. Once speech is recognized, it is then corrected using Language Model. The corrected transcript is then translated to English. Once translation is done, our system will extract CV specific data and  
display it on a professional CV template 
Main Features include: 
• Automatic Urdu speech recognition (ASR) from live audio input 
• Transcript correction using Language Model 
• Urdu text to English text translation  
• Identification of CV specific keywords from translated text 
• CV generation from identified Key Words 
 
 
 
 
 
 
 
 
Technology Used: 
TensorFlow, React, Node.JS, Flask 
Supervisor Name: 
Mr. Umair Arshad 
Group Members:   
Hamza Ali (i15- 0083) 
Talha Mujahid (i16- 0179)                    
Alina Baig (i17- 0327)