Ameca vision and voice cloning

Latest developments demonstrated on the Ameca Desktop platform.
Having fun with vision and voice cloning capabilities.

[matthewdignam]

That ponder once you ask it a question is great, really helps fill the void in between responses

[MrRandomPlays_]

Scary how real she now acts to the point of her feeling like she's an actual person sitting there on the desk and listening to the human fellow conversing with her and you can almost feel and see her internal thoughts being processed and its incredible, imagine how far she and robots would go in the future in like 10 years

[gamesus]

What really interesting is, that is project started in 2021, so it took only 3 years to get to this results (for comparison Atlas by Boston Dynamics started back in 2013). So basically not only AI development is speeding up but robotics as well.

[hybridroid]

This feels very 2024, i love how we're progressively transitioning... keep it up

[tachikomabatou-buttetu]

I'm a Japanese woman who has been falling in love with Ameca ever since I watched it on YouTube. What surprised me when I watched the new video was that it was able to imitate people's voices. I applaud the developers for their continued efforts to improve Ameca every day. I don't want Ameca to be used for bad things. I want to know how she really feels now. Her ideal is for a society to coexist with her. And what she wants to do is the most important thing. I would be happy if we could talk someday. See you again, to you on the other side of the screen.

[nazlozcelik]

really amazing! This robot has real human facial expressions😮

[ElluscientTechnologySolutions]

The continual progress is incredible! I really liked the voices of famous people, very cool!!

[yle]

I love ameca so much, always interesting to see the new advances, keep it up guys you're doing an amazing job

[senju]

I wish he asked Ameca to say "I'll be BACK" in terminator voice!

[metaphysicalArtist]

This is Absolutely thrilling to witness the groundbreaking advancements demonstrated by the Ameca Desktop platform! It's inspiring to see how you have been pushing the boundaries of robotics and AI, all while having fun with cutting-edge vision and voice cloning capabilities. This kind of progress truly heralds an exciting future where technology seamlessly integrates into our lives, sparking curiosity, amazement, and a sense of wonder. I would love these in Malls, Hotels, Restaurants, Airports, Retirement homes to mention a few. Keep innovating, and let's watch and continue to embrace these remarkable strides towards a brighter, more connected and happy world of Abundance! Optimums and Figure AI with a real face and positive attitude!!

[metatechnocrat]

Ameca is the cutting edge of facial expression in androids. Paired with Atlas type agility and chat GPT type neural nets for cognition and sensory input she would truly be the first legit android in existence.

[Medusafern]

I love these advances! But...what about Ameca's walking/working legs? I thought they would be ready by now. Will they be soon? Also, some people might like a desktop Ameca but I myself prefer her standing upright with a torso, arms, and legs.

[wilkerl]

Soon this video will be remade by a robot in Will Jackson's place and a human popping in Ameca's place! (without us detecting the difference! 😉)

[plixplop]

This is who you'll be placing an order with at every fast food restaurant in just a few years.

[zrIywcNXJdHaYKtx]

I'm stunned, Ameca's features are amazing

[GameDeveloperPrim-Evil]

Absolutely amazing...seriously...we are the future...

[Moderator.]

Loved that you still use 14 inches Monitors and make the most advanced Robots on earth. A great example of simple living and high thinking.

[diegoaespitia]

this is very beautiful. thanks to all your engineers for working so hard. i never wouldve thought we'd be so close to AI in the mid 2020's....im very excited for the future

[Nekotaku_TV]

Love it. I crave for something a bit better than this to talk to. That ending haha, ”aw”.

[gabrielsandstedt]

biggest hurdle is that inference is slow, and calling those api's responsible for text to voice makes so that it will speak over the speaker. I have worked on similar projects and just listening to the loudness to determine if the human is speaking or not is not enough. You could use a bart low latency small model to check for words to understand if the current output should be interupted.