SPEECH EMOTION RECOGNITION

<<Supervisor Name>> <<Student Name>>

<<Designation>> Roll No:<<>>
Dept. of CSE Dept. of CSE
LBRCE LBRCE
 OUTLINE
• Introduction
• Objectives
• Related work
• Proposed work
• Methodology

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
2
 Introduction
• Emotion plays a significant role in daily interpersonal human interactions. This is essential to our
rational as well as intelligent decisions. It helps us to match and understand the feelings of others by
conveying our feelings and giving feedback to others. Research has revealed the powerful role that
emotion play in shaping human social interaction. Emotional displays convey considerable information
about the mental state of an individual.

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
3
 Objectives
• Here, the speech emotion recognition is based on the Recurrent Neural
Network (RNN) algorithm which uses different modules for the emotion
recognition and the classifiers are used to differentiate emotions such as
happiness, surprise, anger, neutral state, sadness, etc. The dataset for
the speech emotion recognition system is the speech samples and the
characteristics are extracted from these speech samples using LIBROSA
package. The classification performance is based on extracted
characteristics. Finally we can determine the emotion of speech signal.

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
4
 Related work
• Conventional SER

Early SER studies searched for links between emotions and speech acoustics.
Various low-level acoustic speech parameters, or groups of parameters, were
systematically analyzed to determine correlation with the speaker's emotions. The
analysis applied standard classifiers such as the Support Vector Machine (SVM),
Gaussian Mixture Model (GMM), and shallow Neural Networks (NNs). 

Majority of low-level prosodic and spectral acoustic parameters such as

fundamental frequency, formant frequencies, jitter, shimmer, spectral energy of
speech, and speech rate were found correlated with emotional intensity and
emotional processes.

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
5
 Proposed work

Our Proposed SER system consists of four main steps. First is the
voice sample collection. The second features vector that is formed
by extracting the features. As the next step, we tried to determine
which features are most relevant to differentiate each emotion.
These features are introduced to machine learning classifier(RNN)
for recognition.

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
6
 Methodology

Block Diagram

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
7
 Methodology
Algorithm

• Step 1: The sample audio is provided as input.

• Step 2: The Spectrogram and Waveform is plotted from the audio file.
• Step 3: Using the LIBROSA, a python library we extract the MFCC (Mel
Frequency Cepstral Coefficient)
• Step 4: Remixing the data, dividing it in train and test and there after
constructing a RNN model and its following layers to train the dataset.
• Step 5: Predicting the human voice emotion from that trained data (sample
no. - predicted value - actual value)

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
8
 References
1. Y. Chen, Z. Lin, X. Zhao, S. Member, G. Wang, and Y. Gu, “Deep Learning-Based Classi fi
cation of Hyperspectral Data,” pp. 1–14, 2014.

2. L. Chua and T. Roska, “The CNN Paradigm,” vol. 4, no. 9208, pp. 147–156, 1993.
3. X. Xu, J. Deng, E. Coutinho, C. Wu, and L. Zhao, “Connecting Subspace Learning and
Extreme Learning Machine in Speech Emotion Recognition,” IEEE, vol. XX, no. XX, pp.
1–13, 2018.

4. Z. Huang, J. Epps, D. Joachim, and V. Sethu, “Natural Language Processing Methods

for Acoustic and Landmark Event-based Features in Speech-based Depression
Detection,” IEEE J. Sel. Top. Signal Process., vol. PP, no. c, p. 1, 2019.

5. P. S. Member, “Transfer Linear Subspace Learning for Cross-corpus Speech Emotion

Recognition,” vol. X, no. X, pp. 1–12, 2017.

02/11/2021 Lakireddy Bali Reddy College of

Engineering, Mylavaram
9
 Thank You
02/11/2021 Lakireddy Bali Reddy College of
Engineering, Mylavaram
10