The task of speech emotion recognition in human-computer interaction is both fascinating and difficult. The practice of attempting to discern affective and emotional states in speech is known as speech emotion recognition. Speech emotion recognition is the process of accurately anticipating a person's emotion from their speech. Many states, such as tone, pitch, expression, behaviour, etc., can be used to forecast an individual's emotion. A select few of them are thought to be able to convey emotion through speaking. With the aid of feature extraction approaches that extract characteristics like MFCC (Mel frequency Cepstral Coefficients), chroma, and Mel spectrogram, the proposed system in the project is able to recognize emotions on dataset. Additionally, the suggested system has real-time emotion recognition capabilities that enables possibility to identify emotions.