Tone AI

As AI becomes more integrated into daily life, and consumers communicate more with vocal AI powered interfaces, the need to have intelligent backends to comprehend subliminal human intentions is more important now than it has ever been before. Tone I seek to find a solution to both the growing tensions between human and AI interaction and AI support role comprehension. To achieve this, our team will research just how vocal tones are mapped to emotional intention, create a dataset, design and train a model, and create an API. The output could be used by other AI systems to help identify emotional output. We aim at also adding other features such as noise filtering, example resources on use cases, and research documents to help further research. We are working with multiple faculty at SJSU, and aim to release our findings and products by the Summer of 2026.