• Shreyas Mudaliar
• Mukesh Choudhary
• Abhijeet Iyer
• Ganesh S,

The paper proposes the development of a Speech Recognition system using Mel filter techniques. Speech recognition is used to recognize the speech words faster, efficiently and accurately. Mel Spectrogram is used to find characteristics from spoken word voice sample with respect to the word uttered by the speaker. Convolutional Neural Network (CNN) architecture with Batch Normalization is used and an accuracy of 92-94% for 12 labels is attained. This model can be further improved by EdgeSpeechNet or DeepSpeech and image recognition models.

Speech Recognition, CNN, MFCC, Mel Spectrogram, Batch normalization

"Speech recognition for simple speech commands", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 5, page no.544-549, May-2019, Available :http://www.jetir.org/papers/JETIR1905D79.pdf

"Speech recognition for simple speech commands", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 5, page no. pp544-549, May-2019, Available at : http://www.jetir.org/papers/JETIR1905D79.pdf