Depression Detection of Audio, Image and Text Using Machine Learning Technique

Depression Detection of Audio, Image and Text Using Machine Learning Technique

 Mrs. Shikha Pachouly, Praveenkumar Saw, Pratik Nawale , Gaurav Khaire , Akshay Padwal

Mrs. Shikha Pachouly(Assistant Prof. at AISSMS College of Engineering Pune (M.E IT / PhD Pursuing)

Praveenkumar Saw/Computer Engineering (All India Shri Shivaji Memorial Society’s College of Engineering Pune)

Pratik Nawale/Computer Engineering (All India Shri Shivaji Memorial Society’s College of Engineering Pune)

Gaurav Khaire/Computer Engineering (All India Shri Shivaji Memorial Society’s College of Engineering Pune)

Akshay Padwal/Computer Engineering (All India Shri Shivaji Memorial Society’s College of Engineering Pune)

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - Depression is common mental disorder that affect millions of people in world.in past few years, depression detection has gained popularity through its speech. However, many challenges remain which includes feature providing the best discrimination between classes and depression levels in this model we give different type of features for depression detection by using comparative analyses by using this same collection, the test base and speech base system is evaluated and how the system is built we found that performance can be increase by using combination of feature which is drawn from both speech and text. Nowadays there are lot of platforms like Twitter, Facebook, and Instagram for an individual to express their emotions which can be in the form of images, videos, audios and mostly through text. These social media platforms provide huge quantity of user data which can be used for explorative analysis. This textual data is most widely used data analysis which offers a bunch of characteristics and the state of emotion of an individual. So, we will be using Emotion artificial intelligence for emotion detection, which involves the field of data mining. Twitter, Facebook, YouTube are number of social media platforms. Emotions, opinions of majority people are expressed through their social media images, videos and text. Hence due to this presence of social media there is large user data available for explorative analysis. Mainly textual data is most widely used which gives number of characteristics. Hence it became best choice for data analysis, for emotion AI.

Key Words:  Depression, Data, Audio, Video, Feature extraction.