Our Text to Speech System converts written text to natural-sounding speech based on any voice. The system is constructed entirely from deep neural networks. It is available for any type of voice.
All you need is dataset that contains voice examples with appropriate texts. We need 7 minutes audio samples to create your own voice model.
All speech data was recorded using an identical recording setup: an omni-directional head-mounted microphone (DPA 4035), 96kHz sampling frequency at 24 bits and in a hemi-anechoic chamber of the University of Edinburgh. All recordings were converted into 16 bits, were downsampled to 48 kHz based on STPK, and were manually end-pointed. All speech data was recorded without any emotions.
| Phrase | Generated | Original |
| One season, they might do well. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/1.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/1.mp3 |
| I drove the ball well. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/2.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/2.mp3 |
| We knew nothing about it. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/3.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/3.mp3 |
| They should stop the bombing. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/5.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/5.mp3 |
| I've learned from my mistakes. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/6.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/6.mp3 |
| It became a book by itself. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/7.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/7.mp3 |
| I took the gun. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/8.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/8.mp3 |
| I have no sympathy with her at all. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/9.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/9.mp3 |
| I enjoy the creative process. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/10.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/10.mp3 |
| In each case they were a goal down. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/13.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/13.mp3 |
| It is good for our team. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/14.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/14.mp3 |
| That will be the criteria for the future. | https://ai.arvilab.com/storage/app/media/audio/case1/generated/15.mp3 | https://ai.arvilab.com/storage/app/media/audio/case1/original/15.mp3 |
We also train our system to generate good quality audio from audio dataset with noises. We created our own dataset using phrases from famous series Desperate Housewives.
Check the examples.
| https://ai.arvilab.com/storage/app/media/voice_photos/bree.jpg | https://ai.arvilab.com/storage/app/media/voice_photos/lynet.jpg | https://ai.arvilab.com/storage/app/media/voice_photos/gebriel.jpg | https://ai.arvilab.com/storage/app/media/voice_photos/susan.jpg |
| Bree | Lynette | Gabriel | Susan |
| This is how we train to clone voice. | You miss one hundred percent of the shots you don't take. | You may only succeed if you desire succeeding. You may only fail if you do not mind failing. | Once you choose hope, anything is possible. |
| https://ai.arvilab.com/storage/app/media/audio/case2/bree/1.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/lynet/1.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/gabriel/1.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/susan/1.mp3 |
| Opportunities don't happen. You create them. | The only place where success comes before work is in the dictionary. | Our mind has no limits. | Nothing great was ever achieved without enthusiasm. |
| https://ai.arvilab.com/storage/app/media/audio/case2/bree/2.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/lynet/2.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/gabriel/2.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/susan/2.mp3 |
| Once you choose hope, anything is possible. | A mind that has no purpose, always wanders in the dark. | Keep your face to the sunshine and you can never see the shadow. | We are a team united by the same goal. |
| https://ai.arvilab.com/storage/app/media/audio/case2/bree/3.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/lynet/3.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/gabriel/3.mp3 | https://ai.arvilab.com/storage/app/media/audio/case2/susan/3.mp3 |
This technology can be used for:
- Audiobooks. Create audiobooks with different voices.
- Assistants and chatbots. Create a unique voice for your personal assistant.
- Hotlines. Change robotic voice to the pleasant one.
- Movie dubbing. Movies can be dubbed using the voices of original actors.
- Entertainment. Create and change voice for your character.