Emotion Detection in Speech
In recent years, the research on emotion detection in speech has gained significant attention due to its potential applications in various fields such as healthcare, customer service, and social robotics.
This paper literature review provides a detailed and thorough overview of the current techniques and methodologies including the acoustic features-based, deep learning-based, linguistic features-based, multimodal, ensemble, and transfer learning-based approaches.
Furthermore, this paper discusses the strengths and limitations of each approach. For instance, acoustic features-based approaches are straightforward yet efficient, while deep learning-based approaches can learn complex speech signal representations.
Additionally, linguistic features-based approaches prove useful when the emotional content is closely linked to the speech content. Multimodal approaches integrate information from multiple modalities to enhance accuracy, while ensemble approaches merge multiple classifiers to improve the system’s robustness. Also, transfer learning-based approaches transfer knowledge from related tasks to improve performance in situations where there is limited training data.
The review emphasizes the importance of developing accurate and robust emotion detection systems, which will play a vital role in enhancing human-machine interaction and the success of various applications.
This paper discusses a Machine learning algorithm to discern the emotion associated with human speech developed to account for three emotions namely, normal, angry, and panicked
Sr Data Analyst at Bridgetree
Dishant Banga received his bachelor’s degree in Mechanical and Automation Engineering from Guru Gobind Singh Indraprastha University, Delhi, India in 2014 and Master’s in systems Engineering and Engineering Management with specialization in Data Analytics/ Data Science from the University of North Carolina, Charlotte in 2018.
His interest includes applications of Machine Learning and Data Science to solve complex business problems.
He has participated in various national and international level competitions and secured good rank by developing solutions for complex business problems.
Currently, he is working as a Sr. analyst at Bridgetree, LLC. His research interests include developing statistical and machine learning models and applications, artificial intelligence