Deep Learning-Based Speech Emotion Recognition Using Librosa

Abstract

Speech Emotion Recognition is a challenge of computational paralinguistic and speech processing that tries to identify and classify the emotions expressed in spoken language. The objective is to infer from a speaker's speech patterns, such as prosody, pitch, and rhythm, their emotional state, such as happiness, rage, sadness, or frustration. In the modern world, one of the most crucial marketing tactics is emotion detection. For a person, you might tailor several things in order to best fit their interests. Due to this, we made the decision to work on a project where we could identify a person's emotions based just on their speech, allowing us to handle a variety of AI-related applications. Examples include the ability of call centers to play music during tense exchanges. Another example might be a smart automobile that slows down when someone is scared or furious. In Python, we processed and extracted features from the audio files using the Librosa module. A Python library for audio and music analysis is called Librosa. It offers the fundamental components required to develop systems for retrieving music-related information. Because of this, there is a lot of potential for this kind of application in the market that would help businesses and ensure customer safety.