OTO

Being a student, researching in a similar field of Speech Emotion Recognition/ Behavioural AI, I am really impressed with the product. I tried some labeled use-cases with DeepTone to find out results and compared them with original labels. Here are the things I liked the most about DeepTone.
1. The user interface. I found it really very interactive and easy to use. It raises your interest in the product.
2. Classification of Speech, Music, or Silence works completely fine. The voice model provides the results with high accuracy. 3. Classification of the voice of male or female provides accurate results.
4. The mapping of the energy of the speaker always shows correct results. I feel it gives more understanding of recognizing user behavior or behavior. 5. The best part is emotion recognition. I found it really amazing, which classifies the speech according to user behavior i.e. Happy, Irritated, Neutral, or Tired. I found emotion recognition of the DeepTone voice model works with great precision.

Overall I think, this product can be the best option to find out customer behavior and it can be a really useful tool for reaching out to target customers. I surely recommend this to fellow businesses/customers.

Only one thing concerning I realized, sometimes the DeepTone model gives confusing results in emotions (Specifically Irritated voice and Happy voice) with high-energy tones. I think the doors are open for improvement in this type of scenario and it can also be an area of research. Overall I am really happy with the product.