What is miniTalk™?
The miniTalk™ messenger is a vocal platform for meaningful conversations. Each voice message is limited to 5 seconds, which encourages conciseness. The cool part is that you don’t have to actually say anything to send a voice message. The underlying voiceDNA™ technology analyzes your texts and synthesizes your voice messages automatically. Not with Siri’s voice, not with Alexa’s voice, but with your voice. You can record a voice message the traditional way, and voiceDNA™ transcription will apply automatically. We also want to foster real relationships through meaningful conversations, so you have to receive a sufficient number of messages to fully reveal the other party’s profile. Your results and experiences may vary, but please give it a try and let us know your feedback. Multi-language support is ongoing.
How is miniTalk™ useful?
People have been using the miniTalk™ messenger for:
- personalized text-to-speech (TTS)
- voice messaging
- improving enunciation and pronunciation
- language exchange (English, Chinese, Korean, French, Italian, etc.)
- meeting new friends
- dating, including long-distance relationships
How do I get text-to-speech for my own voice?
There are two ways to get text-to-speech for your own voice: exchanging voice messages with other users when you chat and providing voice samples through the voiceDNA™ trainer. With the latter, you would provide no more than 1 hour of your voice recordings.
To use the voiceDNA™ trainer: Listen to the voice samples and record your own according to the scripted prompts. Pronounce and enunciate as clearly as you can in a quiet setting to get the best results. Once your recordings sound good to voiceDNA™, you can save and continue. After reaching 100%, you can start using text-to-speech with your own voice.
To use the voiceDNA™ trainer: Listen to the voice samples and record your own according to the scripted prompts. Pronounce and enunciate as clearly as you can in a quiet setting to get the best results. Once your recordings sound good to voiceDNA™, you can save and continue. After reaching 100%, you can start using text-to-speech with your own voice.
How can I practice my English?
You can exchange voice messages with new friends and use the voiceDNA™ trainer to practice and test your pronunciation and enunciation. The voiceDNA™ trainer will tell you what it heard and keep track of your progress. More languages including Chinese and Korean are in the pipeline, so stay tuned.
Why can't I see his/her profile picture?
We think other platforms focus too much on pictures and aren’t conducive to meaningful conversations, so with miniTalk™ you can't completely see the other person's picture until you talk with them sufficiently. As you chat with each other and get to know one another, the profile gradually reveals itself. It currently takes 25 messages to fully reveal a profile picture.
How do I make a voice broadcast for my profile?
If a miniTalk™ member has a voice broadcast available, you will see the broadcast icon next to their profile in the users list. Tap on the icon to play their broadcast. To record your own broadcast, tap into your own profile. Record your broadcast by holding on the red button. This page also saves your previous broadcasts. Only the latest recording will be accessible from the users list.
Why use DNA? Is it an hourglass?
DNA symbolizes how we bind and develop together. Both strands of the DNA are connected and resilient. DNA stores information that determines how we are all both unique and alike. Each of our voices is unique, but we communicate in the same languages. From a technological perspective, the voiceDNA™ technology seeks to identify the DNA for each of our voices. It also serves as an hourglass to remind us to communicate concisely.
Who are Talky and Smiley™?
Talky and Smiley™ are the two sides of a conversation connecting like the two strands of the DNA. Talky likes to talk. Smiley is happy to listen. Are you a Talky, or are you a Smiley?
|
What's the vision, what's the future?
As I think about how every aspect of life is being disrupted by Artificial Intelligence and virtualization, voice technologies is one of the most exciting and underdeveloped areas. Why is it that we have to listen to Alexa’s voice or Siri’s voice? Perhaps you prefer to program your doorbell to greet guests with your own voice and wake up to an alarm that is your mom yelling at you. Robot clones of humans would be more complete with voice. Given the latest advancements in Artificial Intelligence and machine learning paradigms, this feat is more effectively achieved than ever.
I’ve developed the miniTalk™ and voiceDNA™ systems to solve the problems necessary to make personalized speech synthesis and voice recognition practical and ubiquitous in daily life. miniTalk™ and voiceDNA™ consist of a vertical stack of technologies, ranging from the web API server, data store and machine learning backend to the client-side mobile app interface. The miniTalk™ messenger is a vocal platform for meaningful conversations and also serves to procure and synthesize human speech. To facilitate frequent and engaging interaction on the platform, all voice messages are limited to 5 seconds and the other party’s profile is gradually revealed as they converse more with you. The fastest way to obtain your personalized speech synthesis, also known as text-to-speech (TTS), is to provide voice samples for scripts we prepared via the voiceDNA™ trainer.
I’m excited about all this because pretty soon you won’t have to actually say anything for your voice to be heard.
-- Art
https://art.wangperawong.com
I’ve developed the miniTalk™ and voiceDNA™ systems to solve the problems necessary to make personalized speech synthesis and voice recognition practical and ubiquitous in daily life. miniTalk™ and voiceDNA™ consist of a vertical stack of technologies, ranging from the web API server, data store and machine learning backend to the client-side mobile app interface. The miniTalk™ messenger is a vocal platform for meaningful conversations and also serves to procure and synthesize human speech. To facilitate frequent and engaging interaction on the platform, all voice messages are limited to 5 seconds and the other party’s profile is gradually revealed as they converse more with you. The fastest way to obtain your personalized speech synthesis, also known as text-to-speech (TTS), is to provide voice samples for scripts we prepared via the voiceDNA™ trainer.
I’m excited about all this because pretty soon you won’t have to actually say anything for your voice to be heard.
-- Art
https://art.wangperawong.com