Speech emotion recognition based on brain and mind emotional learning model

Area of Science:

• Artificial Intelligence
• Cognitive Science
• Signal Processing

Background:

• Speech emotion recognition is crucial for natural human-computer interaction.
• Existing models face challenges in accurately identifying emotions, especially in noisy environments.
• Understanding the brain-mind relationship offers a new avenue for improving emotion recognition.

Purpose of the Study:

• To propose a novel speech emotion recognition model based on brain and mind memory systems.
• To computationally model the interaction between brain short-term memory (BSTM) and mind long-term memory (MLTM) for emotion recognition.
• To evaluate the model's performance and robustness against noise.

Main Methods:

• A two-part model was developed: brain short-term memory (BSTM) for initial processing and mind long-term memory (MLTM) for knowledge storage.
• Emotional speech signals were used as input to the BSTM.
• The model's efficiency was tested by analyzing its performance under varying noise conditions.

Main Results:

• The proposed model demonstrated a strong capability in recognizing human emotions from speech.
• Performance remained robust even when input signals were subjected to different levels of noise.
• Experimental results validated the model's effectiveness compared to existing recognition methods.

Conclusions:

• The novel BSTM-MLTM model offers a promising approach to speech emotion recognition.
• The model's resilience to noise highlights its practical applicability in real-world scenarios.
• This work suggests a new perspective on the brain-mind relationship in information processing.