@@faheemahmadofficial7701 we need someone who can dub in English, Hindi, Telugu. Do you know these languages? If yes then My budget is 120 USD only. Per month full time.
Hindi in thumbnails ultimate way to get clicks ...haha But team the biggest issue for me today is subtitles. Custom models for my speach for hindi to english
I think, the glossary CSV thing can also be configured for subtitles API too, for some level of customisation as per the frequently used words. BTW mai aapki videos dekhta hu, your content is really great 🔥
Someone actually did it, it was IBM or Google I dont remember but they built an AI that could replicate someone's voice by hearing it read a specific line once. They didn't make it publicly available though 😂 I think we all know why. They did it for research and publicly showed results. It was pretty good.
I am using Dubverse to dub videos using AI. The speakers are much better than any AI speaker out there. THey have most Indian languages that no one is doing.
Some language have different sentence for different gender like the one you used (hindi) , it was considering you as a male speaker, so I think they should use voice frequency(pitch) to identify gender in such types of language.
🤣 REKT. I just uploaded an entirely Hindi dub here, though there was 0 human intervention so no promises on the quality! What do you think? ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-0y8uLxTkB3k.html
My suggestion is why not google take this more seriously and a put this in RU-vid as an option. Then RU-vids content will universally be accepted by any country people. This way it will be good for countries with less content in native language and lower prominence of RU-vid.
There are a lot of factors that can not be solved that easy, specially without human input. Also you have to think of two scenarios one where a person has to watch a video and one where a person wants to watch a video. Depending on the end usecase, clientsatisfaction and viewtime may significantly rise or fall.
There’s already an extension that does this so creators don’t have to bother with it! Here’s how anyone can set it up ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-BeS2zKSA6w4.html
Matching the tone and pitch of voice would add so much to the dubbing. I know that at least in English there are various options to choose from. Not sure about other languages though.
When will you release speech to speech translation? You wrote about this technology in Google AI Blog in 2019. ai.googleblog.com/2019/05/introducing-translatotron-end-to-end.html?m=1
Why this kind of effort. You just need is translate app ( from Google ). You speak it and translate it. Click on the speaker button at the bottom. That's it... 😂
The work is really great and explained in quite simplicity... Making it Real time may be a way with some difficulties but it's parameters are nicley tuned👍 thank you for sharing..
That's my dream to have a similar option like changing subtitles to a different language. And, same for the voice, to change the original voice to a couple of different languages in RU-vid.
I need to translate my own You Tube channel from Spanish to portuguese. I don't know nothing about programming. Can you help me? Or I can do that even though I donn't know nothing about programming? Thank you very much. ❤👍 Excellent video. This is just what I need💪👍
Hi Dale, I was thinking about it for a month or so that why not videos can be translated in voice mode. I think you should proceed with this process and brings it over to RU-vid :)
I actually would watch videos with this kind of translation. The future use for this is just amazing! For people who are too young to read or can't read a whole new world opens with this
I have a question, why is not Venezuela in the countries list for making a Google Cloud Platform Account? 😔😔 Is there a video of how to use Google Cloud Platform? Billing, Credentials, Api key, auth, with a step by step?
Combine this with ai voice cloning like lyrebird and you can speak any language in no time... That makes something fascinating and scary at the same time
I don't understand why RU-vid isn't doing this already, so much content could be made available in native languages. Lots of comments here are asking for RU-vid to do this. Do it RU-vid!
Hey, good work with the video. Very interesting. However, if you listen to how you communicate with your voice, you will notice that a lot of the communication is lost when AI steps in. AI has strange nuances and undertones (and a great lack of other). Human communication using the voice goes much further than the words spoken. I would watch a short AI dubbed video as an experiment, but I think I would get bored and estranged by it. I would much rather watch a video in the original language with subtitles (even the crazy and flawed auto-generated ones) than listen to the AI rendering of your speech in another language. I fear that, in the future, dubbing, even in studios, might be performed by machines instead of actors. My fear is for the human psyche and humanity's soul. If people get used to this dehumanized speech, more and more communication and emotional problems will rise. If you grow up listening mainly to this, as you would if AI takes over dubbing and you grow up with television, computers, and so on, you will have a hard time expressing and recognizing emotions, you will become less empathetic. You become colder, more distant and, eventually, humanity will loose a big chunk of it's humanity. AI seems to be a great tool that could free peoples time (lots of jobs are being replaced which could mean more free time for people or people starving depending on how governments deal with this). However, when it comes to art and communication I deeply believe it shouldn't be used as a way to replace the artist. Perhaps as an aid for some of the technical areas, but not as a replacement of the one doing the expressing of feeling and ideas, nor even the one making the directing and final editing. Have you seen AI generated images? When it come to creating images form scratch, even copying the style of human artists whose works have been fed to it, it is always lacking the essential. The same happens with the voice. Cheers.
Nice!!! This is amazing! Great job! Im working on creating a machine learning model that trains with previous dubbed episodes and dubs future episodes with each character unique voice.
"Hey google" Are you guys leonardo DiCaprio from Inception movie... Because u steal my idea... Yeah.. It was my future solo project to work on.. Btw.. Interesting
I am really wishing this is able now : Any updates on google translation capabilities? Say if one had a dance course can it hear .see what you are saying and dubbing into multi languages and give subtitles to each new language it produces? Would be really great if this exist now with all new AI capabilities ? Also if the translation would be able to change if it’s a woman talking or man in the same course to their audience
Hey google :) Would it be possible to mixe that technique with generatives networks to have lips synch by altering the video feed? Like Károly Zsolnai-Fehér of two minutes paper, couple paper down the road I can see a pipeline for automated video translation close to VO feel.
imagine how cool it would be if in future we can watch netflix video from another language to our native language even if it is not available but just using subtitle and Realtime video dubbing using AI
@@daleonai7448 Yes, it does, but it doesn't dub the video. Would like to eventually be able to get dubbed translations for videos with a similar voice.
When restricted to using commodity services (e.g. google APIs), this is all good info. However, results from state of the art research is way above the quality of commodity services.
Early on I contemplated using ML to build a model that matches actor's voices to machine-made voices from the Google Cloud Text-to-Speech API. I would build a vision model using AutoML Vision that compares the mel spectrogram of voice samples. But I ran out of time. You should try it!
