I used to think ML would be daunting and boring. Instead, Dale’s enthusiasm and daily life applications of ML really illustrate that this can be fun and approachable. 😲
Wow.. the first use case 'make multimedia searchable', I have discussed it with colleagues many times. I just didn't know that the solution is already almost there.
One of the most beautiful advertisement I saw today. Thought it was about ML and not about selling Google products (and giving him the few spare things don't have yet)
Really amazed by Google Machine learning tools and your presentation skills👏👏. I'm looking forward to learn Google ML tools. Edit: That Hindi translation was too short to be judged for its accuracy, but the idea and the translated voice was quite impressive though. Google translated English text is not always that accurate. It has improved a lot though .
This is so cool - it really illustrates a lot of key topics in ways that are useful and relatable. Thanks for sharing! You've also come along way from those first steps! ;)
What is "cool"? That software can detect bicycle in video and mark that second minute contains it? Or possible to find voice "hi dady"? Oh my, people who u are to be fascinated by such things... What's really shocking is that u believe this is "artificial intelligence"...
Is all this stuff available for the public or do you have to be a google developer, to have access. I wonder if you would do a walkthough of one of your demos, showing exactly how it's put together.
Great video and nice hindi translation. What kind of mic you are using your voice is fantastic and very clear. These are great example for ML. Thanks for the video.
2:30 Sparkly biiiiike! :-) I agree that finding such moments is an excellent use case. My barrier right now is the cost of encoding & cloud storage, but I look forward to searching through old home videos (and film) someday.
Wow, amazing video, very inspiring! I liked how you explained how to use Machine Learning to solve real world problems 🤩I like the approach you did and the projects using Google Cloud Tools, everything amazing! PD: It's so nice you have a lot of recordings of your childhood and now you cand find them easily 😁
I have ZERO knowledge of ML and AI but found this video interesting. It may sound like a dumb question but what if videos don't have external tags, transcripts, and descriptions with them. How would you make sense of them using ML & AI tools? Thank you
The example in the video used a ML model from google. So it's basically just loading in the video, let the model do it's calculations and take the output. What the model does under the hood is very comlicated math - and also a lot nobody actually knows. In this example, it used some voice recognition to transcribe speech, propably image-recognition to identify certain objects and areas and then the word embedding to feed the search-engine. The word embedding is seperate because the words you write into the search are send through the same embedding algorithm - but obviously you wouldn't send those words through the image-recognition. Image recognition might be done on every image or once per second or maybe it identified a scale of "change" so that if the camera was filming the same spot for a minute, the model does only go over it once instead of 60 or so times. The reason this works is because Google used billions of images which were already tagged, have ML and AI scientists create a theoretical construct that should be able to learn identifying those -> and then have that construct do a massive ton of calculations and self-adjustments (the actual learning process) until it found the same tags in the images as provided. This training process might easily have taken days or even weeks. At the end, the model has "learned" to identify these things and can be used on data it has never seen before and hopefully find fitting tags as well. Just to reiterate, nobody knows what calculations the model is actually doing or why they work. Because during training it has finetuned millions, if not billions of parameters for calculations.
Pulling us with the cute baby videos and then amazing us with futuristic tools that make AI accessible... it was a good plan, and it worked on me! Interesting variety of tools here, but the video managed to flow smoothly from one to the other and give a taste for what's available in the API. 👍
All that can be done locally, but she's advocating the benefits of using cloud and google ML APIs in particular. I personally love to apply ML to my daily life tasks to make it better and easier. For example the file searching example, I've made that so that it would scan the locally stored files, since that would be faster and given the bad rep that Google has regarding user privacy, I can't chance uploading my personal files and media to google drive, also it'd be a time and space consuming task.
Awsome, I can now finally understand my boss's unorganized stock system that he used up to 10 different descriptions for the same item!! RUclips algorithm is doing a fantastic job this time.
Nice video which deserves a like....but Dale you have one privilege which I - as a common man - don’t have..which is ..you don’t have to pay for GCP services coz you work there 😊 …even the free tier does not help much-tell your bosses to make it affordable …
0:22 "Machine learning is this completely new thing of getting computers to do things." Whatt??!!?? The term itself was coined in 1959. Linear regression dates from the 1890s, and PCA was introduced in 1901. Machine learning is older than modern computers!
A lot of your viewers are from India and you don't speak a lick of Hindi? Don't worry. Many Indians don't either. You should try making ML algos to try other Indian languages like Malayalam, Tamil, Bengali and so on :)
2:30 - finding specific memories in videos
6:19 - semantic reactor - google sheets
7:31 - Universal sentence encoder
I used to think ML would be daunting and boring. Instead, Dale’s enthusiasm and daily life applications of ML really illustrate that this can be fun and approachable. 😲
step one: share your entire life with google
step two: do stuff
Google: who has their family videos since they were born?
Dale: me me me !!
google: ok you're doing our next TensorFlow video.
Ok Google. 😆
hahahha funny
Well, yah. Big yah.
The information density of this video is excellent! Really liked how you squeezed in so much into a relatively short video.
Very inspiring Dale! Thank you so much. I'm new in this ML world and it makes me wants to do more with it
Use cases are of course great but I love this person the most
😍
Totaly agree! Great communication skills.
The elbow pose over time in the video is pretty cool.
What a great exposure to understanding how to start using and thinking about problems with machine learning. Definitely going to check it out!
AI newbie here. Wow. Just wow! You had me since the AI read the icing on the cake.
Amazing, inspiring and totally relatable examples....
Wow.. the first use case 'make multimedia searchable', I have discussed it with colleagues many times. I just didn't know that the solution is already almost there.
One of the most beautiful advertisement I saw today.
Thought it was about ML and not about selling Google products (and giving him the few spare things don't have yet)
Hindi version sounded pretty good to me. All the examples are pretty amazing
Very useful content fluently presented in a clear and personable style. (I'll be trying the video translation.)
I just loved every bit of this video.
Wow!
I don't know much about ML, and this is really inspired me to know more about AI and ML.
Thanks ❤️
It's a deadly road be tough😂 n don't give up
exactly! same for me
Really amazed by Google Machine learning tools and your presentation skills👏👏. I'm looking forward to learn Google ML tools.
Edit: That Hindi translation was too short to be judged for its accuracy, but the idea and the translated voice was quite impressive though. Google translated English text is not always that accurate. It has improved a lot though .
This is so cool - it really illustrates a lot of key topics in ways that are useful and relatable. Thanks for sharing! You've also come along way from those first steps! ;)
What is "cool"? That software can detect bicycle in video and mark that second minute contains it? Or possible to find voice "hi dady"? Oh my, people who u are to be fascinated by such things...
What's really shocking is that u believe this is "artificial intelligence"...
Eye opener introduction to TF application.
Watching your tennis serve example was really helpful. This could be also applied to other sports too...like say shooting a basketball?
Great preso.Would be greater if you could give some idea around cost for the stuffs built
How much was approximate cost according to you?
You actually solved a serious problem. Thank you so much!
This is a great talk and a great speaker!
I used python to convert text into speech ..once in a tiny project ..and it worked well .. that was easy too !!
I don't know the initial speech but the Hindi translation was not *that* bad! Did the job pretty well
PHEW
Well I think this is best video so far I saw on why so much fuzz out there about ML.
Amazing! Thank you Dale!
How are you tracking position of your skeleton, you wrist, elbows, shoulder etc. in a video?
Great demo!! Thanks for sharing
What a great presentation! This is blowing my mind. I feel like a kid at a candy store. So many possibilities! Thanks!
Is all this stuff available for the public or do you have to be a google developer, to have access. I wonder if you would do a walkthough of one of your demos, showing exactly how it's put together.
Great video and nice hindi translation. What kind of mic you are using your voice is fantastic and very clear. These are great example for ML. Thanks for the video.
As always, a great and inspiring video! Thanks a lot
Wow... this is one of those ideas that I actually want to use
Good Stuff!! The Hindi part was also very convincing.
2:30 Sparkly biiiiike! :-) I agree that finding such moments is an excellent use case. My barrier right now is the cost of encoding & cloud storage, but I look forward to searching through old home videos (and film) someday.
Great video on machine learning application and very relatable
Really interesting video Dale. fun to see what options google and Tensor has.
Very cool. Great idea about the applications of ML.
Wow, fantastic, thank you!
💥
Very awesome video !!
This was awesome! Thanks
Loved it!!
I loved loved loved this video!
The Hindi part made me hit the Like Button 😊
Just wanted to know which technique she might have used to convert text to vector..? Is it word 2 vec?
Probably. That is usually a best use case
Very nice video.Thank you!
Wow, amazing video, very inspiring! I liked how you explained how to use Machine Learning to solve real world problems 🤩I like the approach you did and the projects using Google Cloud Tools, everything amazing!
PD: It's so nice you have a lot of recordings of your childhood and now you cand find them easily 😁
Really helpful..Thanks!
Great talk Dale.
wow! great session. thankyou
That hindi language generated video was really amazing 👍🏼. I think that was even better then what we speak here 😂
But the meaning was lost in translation.
Such nice explanation in such easy language !!
I think supercomputer is required to train this model??, Can anyone
Better use isolated virtual environment when testing the A.I solutions.
Where is all the repository stored, it’s it in key value format ?
Purely inspiring! Dale Markowitz you should start your own RUclips channel!
She explaining everything so cool 🔥Dale Markowitz 💚
8,000
I have ZERO knowledge of ML and AI but found this video interesting. It may sound like a dumb question but what if videos don't have external tags, transcripts, and descriptions with them. How would you make sense of them using ML & AI tools? Thank you
The example in the video used a ML model from google. So it's basically just loading in the video, let the model do it's calculations and take the output. What the model does under the hood is very comlicated math - and also a lot nobody actually knows.
In this example, it used some voice recognition to transcribe speech, propably image-recognition to identify certain objects and areas and then the word embedding to feed the search-engine. The word embedding is seperate because the words you write into the search are send through the same embedding algorithm - but obviously you wouldn't send those words through the image-recognition.
Image recognition might be done on every image or once per second or maybe it identified a scale of "change" so that if the camera was filming the same spot for a minute, the model does only go over it once instead of 60 or so times.
The reason this works is because Google used billions of images which were already tagged, have ML and AI scientists create a theoretical construct that should be able to learn identifying those -> and then have that construct do a massive ton of calculations and self-adjustments (the actual learning process) until it found the same tags in the images as provided. This training process might easily have taken days or even weeks. At the end, the model has "learned" to identify these things and can be used on data it has never seen before and hopefully find fitting tags as well.
Just to reiterate, nobody knows what calculations the model is actually doing or why they work. Because during training it has finetuned millions, if not billions of parameters for calculations.
good to see 'apple' and 'garbage can' is so close in embedding space chart
This is such a good video!
Nobody:
Google: Now I can relive those moments too
Do you have an estimate of how much it would have cost you on Google cloud for all that?
Pulling us with the cute baby videos and then amazing us with futuristic tools that make AI accessible... it was a good plan, and it worked on me! Interesting variety of tools here, but the video managed to flow smoothly from one to the other and give a taste for what's available in the API. 👍
How much do they charge for translation api.
I liked this video by Dale. the presentation is simple and excellent. I feel it engaging and immersive. though I m non tech background.
We all have a excerpt from our life to associate with why we started using ML.
Great video !!
Can we use this as projects Of Machine learning
How can I learn algorithms to do myself? I don't want to use ready-made service, Instead, I want to learn to develop on my own.
did you level the videos
Fun ML stuff 👍🏼😊👍🏼
How long did this app take & how many people worked on it?
amazing video
thank you i can see it now . whereto use Ai but i need to learn it
This was great, but what about doing ML for free, without the cloud?
All that can be done locally, but she's advocating the benefits of using cloud and google ML APIs in particular. I personally love to apply ML to my daily life tasks to make it better and easier. For example the file searching example, I've made that so that it would scan the locally stored files, since that would be faster and given the bad rep that Google has regarding user privacy, I can't chance uploading my personal files and media to google drive, also it'd be a time and space consuming task.
This is great, I love your videos, could you also include info around what the costs would be?
Great Work.. I've left earth..thanks for the cute family videos..
How to convert conversational video to text by each person talking in video?
is Vertex better than google colab or anaconda?
awesome. thank you so much.
Loved it.
I only know the Forrest tree diagrams.
Awsome, I can now finally understand my boss's unorganized stock system that he used up to 10 different descriptions for the same item!! RUclips algorithm is doing a fantastic job this time.
hunger is everyday problem.. can it be solved?
This is a great vid
Great use cases. Thanks Dale
Nice video which deserves a like....but Dale you have one privilege which I - as a common man - don’t have..which is ..you don’t have to pay for GCP services coz you work there 😊 …even the free tier does not help much-tell your bosses to make it affordable …
I mean it's pretty cheap for the most part...
0:22 "Machine learning is this completely new thing of getting computers to do things."
Whatt??!!?? The term itself was coined in 1959. Linear regression dates from the 1890s, and PCA was introduced in 1901. Machine learning is older than modern computers!
Not only google knows everything you after your account created. It is now getting info of you before google was established.
Woah... There are a lots of readymade tools in Google cloud. Are the tools in Google cloud free or I should pay for it?
They're pay-for, but most have very generous free tiers!
@@dalemarkowitz8021 And its amazing that all these things u and Torry have showed 2 years back itself in I/O 19.
@@NaveenKumar-lb5cx Wait, WHAT? This is unacceptable! I'm ashamed!
Amazing!
A lot of your viewers are from India and you don't speak a lick of Hindi? Don't worry. Many Indians don't either. You should try making ML algos to try other Indian languages like Malayalam, Tamil, Bengali and so on :)
Wow I like her😌
imagine handing over hundreds of hours worth of your childhood videos to google.
Thank you far sharing! So personal, So Cool!))) My applications are so less important in terms of meaning of life.
Very inspiring
very nice video :)
Very insightful. Thank you :)