Recent Trends in Speech Processing for Speaker Emotion Recognition-A survey

Abstract

Over the past few decades, Speech Processing technology has been one of the technologies that
is being practiced by wide range of engineers, scientists, linguists. The dominant features fetched from the
speech data always served as key component of decision making. It is therefore, become imperative for
computers or systems to have the capability to understand and decipher human speech. This would help
them resolve decision related problems efficiently, but the task of quantifying the speech features,
channeling them into well distinguishable partitions or states and using them for the purpose of extracting
meaningful information so as to study the different characteristics of cognitive state of human
consciousness has been one of most important challenges for language and psychology researchers. For the
past few years, noticeable amount of research has been done in the field of Linguistic Speech Processing
such as Emotion or Sentiment Recognition. Deep Learning has become emerging technology which is being
evolved as a sub-area of Machine Learning. Deep Learning uses Convolution Neural Network algorithms
that imitate human’s cognitive model of brain and processes uncountable and unsupervised form of data.
This paper aims at detailed understanding of speech data processing using Machine Learning techniques by
exploring the theoretical and practical aspects of the Research problem.