• Pranali Deshmukh
• Pooja More
• Rutika Ovhal
• Mokshada Mahajan

Speech is the way that humans can interact with each other also interact with computers, and understanding speech is one of the most important processes that humans communicate. SER's main goal is recognizing the emotion of humans in a given speech. When a human is speaking then analyzing his/her tone and pitch to underlying emotions while using his /her reflected voice. For that, we can use CNN for more efficient results. We use the MFCC for the extraction of features from audio or given extracted input. We use the dataset for evaluating our models and functions. The evaluations of emotions like Happy, sad, angry, neutral, surprised, disgust of speech will defect or we can find it. And using the same we record or predict the ratings of any product bought on the web app which we used to do earlier manually.

Input Extraction, CNN, MFCC, Chroma, Mel Spectrogram Frequency, Librosa, Sound file, NumPy, Scikit-learn, PyAudio, MongoDB, MLP.

"Speech Emotion Recognition using Convolutional Neural Network (CNN)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 7, page no.a187-a194, July-2021, Available :http://www.jetir.org/papers/JETIR2107027.pdf

"Speech Emotion Recognition using Convolutional Neural Network (CNN)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 7, page no. ppa187-a194, July-2021, Available at : http://www.jetir.org/papers/JETIR2107027.pdf