 Hello, very good evening to all the students, teachers and parents who have joined us. We will be starting just in few minutes. We are very welcome on the day 3 of celebration Scratch week with Pictoblocks. Yes, hi Uday, hi Kunj, welcome all of you. Welcome to the day 3 Scratch week celebration with Pictoblocks. Hi pal, hi Raksana. I welcome all of you to the day 3 of Scratch week celebration with Pictoblocks. Hello Nitin, hello others, welcome all. So Chinmay, your question is like on which website you need to go. You need to go on pictoblocks.ai. On that website you can just, that is a web app of our Pictoblocks. So like a web version, you can go over there and perform an activity. I have to take, it's a very pleasure to see all of you here. So students, we will be starting in just few minutes till the time I would request you all if you have downloaded the Pictoblocks in your laptop, PCs and Android phone. It's wonderful. If not, please download it from the links I am sharing you. Also do not forget to go and visit our web version of Pictoblocks. Yes Pratik, we are going to share the link. Yes, see Ayush sir has already shared the link over there. So for Shivam is asking for repeat the course of yesterday. So Shivam, we are sharing the recorded video for you and it's on YouTube. So you can always go and revise the session which we have taken. I shared you the link for yesterday also. Yes, you can make a multiplayer game in Pictoblocks also. So as you can see in the chat, Ayush sir has shared a link with you. You can go on that link and you can find the replay video of the previous session. So I think what I am going to do is we will wait for one more minute and I will just disable the chat and float out all the links so that everyone can get those links again. Exactly. So this is the link for downloading the Pictoblocks in your laptop. And next up, link for the Google Play Store to download it in your Android phones. Link to the very new web version of the Pictoblocks. And let me also share the link for the last two days replay video. Yes, so here on the chat you are having all the things. Ayush sir has shared all the links important links with you. So you guys can just follow up the slings. Okay. Yes. So and I hope that all of you had tried that artificial intelligence session things like the whatever the codes we have make yesterday you have tried that things and learn something new. Yes. Okay. So I think we should just start over today's topic. Yeah. Yeah, Charlie. I think we are good to go. So students, I'm enabling your chart. If you find any trouble, please let us know. Yeah. Okay. Okay. So I hope like my audio and the video is properly visible and audible to you. Right. If it is, can you please show me the thumbs up so that I can just start our topic. Okay. Great. Great. So like in yesterday's part we have learned about an artificial intelligence and we all have got to know about how good the artificial intelligence is. Right. We have learned how we can make a face recognition based on face expression detection. Right. So now in today's part we are going to learn about machine learning. Right. So I'm sharing my screen with you and let's us learn about machine learning now. Okay. Just a second. Yeah. Is my screen is like visible? Yes or no? Yes. Okay. Great. Great. So let's begin this beautiful journey of machine learning. Okay. Yeah. So in this part in our today's session we are going to get to know about what exactly the machine learning is machine learning process and we are today going to make a very wonderful activity of a face mask detector where we are going to check whether the person has wear the mask or not. Right. And preparing for the competition. So first of all let's start with machine learning. So how human learns as you know we all are humans and we how actually we learns that is a very tricky part. No. Okay. So first basically human learns from four mistakes. Right. And the mistakes are like human learns for the four steps. The first is sense environment. Second is analyze information. Third is decide an act and fourth is increase the knowledge. But like what exactly these four steps are. Suppose just take an example if you go to solve your math examination paper. Right. So first of all what you do you just see the questions. Right. Then you just analyze the questions like what exactly you need to do over here. Then you just decide what answer you should put like you just perform all the calculation and decide what answer you should put. And suppose if you fail like to give the proper answer or if you missed some steps and when you came home you realize those silly mistakes. Yeah. I had just forgot that steps. Right. So what you do for the next time when you go for the math paper you remember those mistakes and learn from those mistakes. Right. It means basically you are just increasing your knowledge. Right. So this is the four steps how the human learns. There is a very well known like a human learns from the mistakes. The mistakes makes the man perfect. Right. Because the human learning itself says like the human will going to learn the things when it will going to perform the mistakes. Right. So I hope this four that these four steps are properly understood by you. The first is sense environment. Second is analyze information. Third is decide an act and fourth is increase the knowledge. Okay. Yeah. Great. So now we are going to like from where human going to learn this thing. So our first step is sensing the environment. Right. So how human sense the environment. You all know like a human have the five sensors. Eyes, ears, they like tongue like your body body from where you can touch the things and your nose from where you can smell the things. This all the five senses perform perform the activity and the things are getting hit with the brain and the brain will going to give you the reaction like how you need to react. Just suppose if there is a fire and if you try to touch the fire. Right. So you are touching with the hand. So suddenly you just pull up your hand. How because your brains give you the command to fill up the hand. Right. So basically these are the five senses from where you can sense the environment. Right. But why it is required for because see once we perform once we sense the things. No. Yeah. Once we perform the sense the things then this sensing related things is basically getting passed for analyzing the information. This sensing related things are getting passed for analyze the information and once we analyze the things. Right. Like if it is a fire we have analyzed the things. It is a fire. So what you will do you will just pull up your hand. Right. So you have decided something like what you need to do. Right. You have decided to pull up your hand. If you have not pulled up your hand basically you will get born. Right. You will get high. Right. So you decided to pull up your hand and when the next time when someone will going to tell you to go and touch the fire will you really going to touch the fire. Not at all. Because you know you have now an idea like a fire is a dangerous thing. So this is the way how actually the human learns. Okay. Great. So now machine learning what exactly the machine learning is. So machine learning is the process of a machine's learn how to act by themselves without any human intervention. Right. Here what happened machine tried to learn by itself. We are like how the human learns it tried to learn by itself only know in the same way machine also tries to learn by itself do the activities by itself and complete the task. Okay. But in the between whole stage of human intervention is not required but at an initial stage that is a starting stage only the human intervention is required for making the things then later on the things works automatically. Right. Just just take an example like you know Alexa we speak something and according to that Alexa will give you the response. Right. So basically the human intervention is required while making an Alexa but later on there is a new human intervention is required. All the things are working automatically. Right. So machine learning is the process of a machine's learn how to act by themselves without any human intervention. It is basically getting a computer to perform a task without explicitly programmed to do so. Right. So basically here the computer will going to perform a task without any human intervention and basically we are going to perform the program at an initial level and later on all the things will going to work automatically. This is what the machine learning is. Okay. So but do we require to give some input to the machines. Yes. At an initial level as I told you earlier initial level the human intervention is required but later on no human intervention is required the things will going to perform automatically only. Yeah. So here you can see input model and output. So what does that means we are going to make a model just to take an example. I have made a model of a dog versus cat classifier. So what I need to do over here. I want to make a model in the way like whenever I will show the image of a cat it will going to show me. Yeah. This is a cat and whenever I'm going to show the image of a dog it will say yeah this is a dog. Right. So but before that the machine doesn't know what does the cat mean what does the dogs means. So we have trained it and in the model we have given the input and it will going to give you the output. Okay. Now the main and important part of what machine machine learning there are two important parts like steps in the machine learning. The first is a training and second is testing. But what does the training and the testing exactly means. So training means like before exam for the past five to six months you do the lot of studies. Right. You basically trained yourself for the exam. You do the lots of like or do the all time study performs all the quotient perform all the mathematical calculations right. So basically you are getting trained yourself. Right. And what is testing the examination time the examination you just give your test to know and based on test you will get your marks. Right. So in the same way it happens in the machine learning to attend initial level we give the lots of an images to the machine so that based on that it will going to perform training. Okay. And after it performed the training we will test it. Okay. And in the testing part it will going to show you whether your model is working fine or not. So in the input image we will going to give some input image of a dog and a cat and in the model that is our program it from our input image will go throughout the model and it will going to give the output like if I give the image of a cat then it will going to give me the output as yes it is a cat and if I give the images of a dog then it will going to give me the output as yeah it is a dog. Okay. So is the machine learning concept is clear to you have you understand this slide. If yes please just to show me the thumbs up. In the chat. Okay. Great. Yeah. So see as I told you the example of a dog cat and all right. So here you can see I have given the image of a cat as an input. Right. And in the model I have made my program and yes as I made my program the cat image will going into that my program and accordingly it will going to give me the output and here the output is cat. But suppose if I give the image of a dog will my model will give me the output as a cat. Not at all because we have performed a great training and as we have performed the great training it will be going to give as an input as a dog only. Okay. Yeah. So like as I told you like we can make the machine learning models in a different way like it is not necessary that you can make the dog and the cat classifier only you can make a different things you can make a like anything like if you want to if you are if you want to make the dog breeds classifier so you can make that too like you can give the different images of a dog's like of a different breeds the lab red or the German Shepherd right the different different breeds you can give and based on that you give an image it will going to give you the output like which kind of a breed it is right isn't it so amazing in the machine learning we can perform a multiple things the more you explore you will going to learn a lot. Yeah. Yeah. So with the help of the machine learning what you can do you can also make a different different games like it is not necessary we can just going to classify the animals only no not at all the more you explore the more you learn like here we can do the lot of things here like you can see the rock paper scissor game so with the help of machine learning you can also the games like the rock paper scissor game where you will show with the hand as a rock the symbol of a rock the paper and the scissor like this and it will going to predict your move like whatever the move you played whether it is rock paper scissor and accordingly the computer will going to play its own move and once it's played its move then it will going to show you the winner like who is the winner like whether you can be the winner like the player or the computer can be the winner this is a very wonderful project yes and with the help of this project you can make a really wonderful different projects too yes so please try to explore the vector blocks and there you can have a learn the machine learning different things also yeah so let's make a mask detector so what does the mask detector means like if I have wear the mask then this it will say please wear the mask and if I am not wearing the mask what it will say like if I am wear the mask it will say like thanks for wearing the mask and if I am not wear the mask I will say please wear the mask as you know in this pandemic situation the mask is our essential need right so we can make such a really good project with the vector blocks okay yes so don't worry like a path don't worry we will send all the links to you again okay so don't worry about it all the links will be going to send to you but before it just pay your attention over here only because here we are going to learn a different things now I am going to make an activity of a face mask detector okay yeah and after making an activity we are going to test it how exactly it works okay so first of all what I am going to do I will go to the vector blocks okay yeah so I am sharing my screen of vector blocks so you can visit an online web version with the help of pictoblogs.ai I hope my screen is properly visible to all you can put your like response in the chat are you able to see it yes we can see it okay okay that sounds great yeah good so now let's make a face mask detector project so first of all what we need to do we need to go to the add extension here we have a multiple different extensions right so if you feel like the blocks given over there in the pictoblogs is not sufficient so you can go to the add extension and how you can explore the different extension there you can find a multiple different blocks so like here I will choose the machine learning okay but here you can see only two blocks are there so as I already told you like we in the machine learning we can perform the two things learning and testing right so in the training part we are going to train our model and in the testing part we are going to test our model so for training we will go to the creator model and for testing we are going to load a model so I am going to click on create a model so once I click on create a model it will directly redirect you to the teachable machine page okay so on this page basically you can perform your training okay so just put your eyes on the screen because this is an important part yes okay so in the machine learning we can train our project with the three different things that can be image, audio and pose but for now we are going to work with the image project only okay so I will go on the image project okay so now this is the exact interface so now what you need to do first of all here you can see the classes yeah the different different classes so now what you need to do you need to basically change the name of the classes how just click over here and I will just put the name of a class like mask okay in the class 2 I will be going to give the name as unmask okay like this so now what you can do here is the option of a webcam and upload like if you are having the images in your computer or your drive so you can directly upload it but we can use the webcam also so I am using the webcam so here you can see my full face is visible over here right so now what I will going to do I will wear the mask okay so just wait for a while like this so I am going to record the number of images in different different directions and in different different way so I have taken up the sample around 180 images right now in the same way I am going to do for unmask okay so in the same way also I am going to record the multiple images in each and every direction just try to do it because the more the image you will give you will receive a good output okay so after doing the after performing the just recording all the images now what you need to do just we need to go on train model okay so once you click on train a model will going to start its training automatically but it will take a some time okay so just hold for 1 or 2 seconds it will going to start automatically okay so like when like here what like you know the people needs to when like when we are a children we don't know about the things know like what is a cat what is a dog right so people shows us the images of a dog and a cat and all these things then at the end now any image of a dog and a cat will come in front of you so you are able to recognize it in the same way it happens over here so now in the preview option you can see as I have not wear the mask it is just saying me unmask right it is predicting it as 100% so now what I will try to do I will try to put up the mask so you can see as a mask even the mask image is coming like a small image is coming like this so it is just appearing it as a mask right yeah so here in the preview option you can check your model whether it is working fine or not if it is not working fine then you can just recapture the images you can recreate the class you can delete the class and upload the things again right so that in the preview you can check whether it is working fine or not so once you have done that don't forget to just turn off the input ok then click on export the model because the training is completed now now we need to perform the testing so we are going to export our model so click on export my model then a pop up screen will going to be appear like this ok then once this screen will going to come what you what you need to do you basically need to go on upload my model ok so your model will going to upload and a link will going to be appear over here see the shareable link is available and you can copy that link right the link is available over here so just click on copy ok so now you can go on your pictoblocks so here what I am going to do now as I go on my pictoblocks so I already told you create a model is for training purpose and load a model is for testing purpose so I am going to go on load my model and here what I need to do copy that link the link we have just copy from the teachable machine know so we need to paste that link over here so I will going to paste it choose my model as a image and load my model yes yes right so we have our model has been loaded successfully so how will you get to know your models has been loaded or not like here you can see image model and all the blocks related to your model is appearing over here you can see mask unmask over here right ok so now let's make a program so I will just put a toby over here for a while yeah so first of all use when flag click ok so what is the use of when flag click I think you already know but basically I will tell you again so whatever the code we write below when flag click will going to execute whenever we click on this green flag so now what I need to do I need to turn on video on a stage with 0% transparency ok so after that what we are going to do we are going to check the things ok so I am going to apply if else ok so in the if part what I will going to put is see this identify class from the web camera is mask ok if from the web camera if it has identified the class as a mask then it will say thank you for wearing mask ok and in the else part what you can do if it is if the person has not wear the mask it will say please wear the mask ok now what we want we want this things into the continuous way you know like it will going to recognize one time and then it will stop so what we can do we can apply the forever block over here like this ok so let's try to run the code and check whether it works or not so it is saying me please wear the mask please wear the mask right because I have not wear it so now what I am going to do just take my mask and put it up here so it is saying thank you for wearing the mask thank you for wearing the mask when I remove it it will say please wear the mask so have you seen how easy it is like once we perform the training perfectly then all the things are very much easy right what you just need to do just remember the things how we export a model how we just copy that link and just paste it into the load model I tell you again like machine learning works into two frame training and testing create a model is for training and load a model is for testing once you once you just perform the training then you need to copy the link from there and paste into the machine learning and the load my model right and once it is get load you need to write your model like your program over here right so in the first part we have seen no like input model and output so here I am going to give the input image of mine then the model is just recognizing all these things and giving me the output like yeah you have wear the mask or you have not wear the mask right so I hope you have enjoyed these things yes if yes please just show me the videos up in the chat so we have successfully just test our project completely right we have successfully test our project completely okay so now let's not run now what is the post detection like when you have just when we are going to the teachable machine the third option you have seen there is a pose the pose detection so the way in the same way we have made our face mask detector using an image detection now in the same way we can work with the pose detection in the same way we can work with the pose detection okay so in the pose detections what happen train the computer to identify different human pose and make interactive projects okay like here you can see in the first image you can see the baseball player is just throwing a ball and what are the dots and the lines so in the pose project how the pose project is basically different from the image project in the image there is no lines and dots it is totally capturing the images but in the pose project it will find out the dots and the lines the dots are basically the joints and the lines are basically body parts like a hand legs right all these things are basically it will connect that joints with the that line and will going to make a perfect pose and with the help of that pose you can identify the multiple you can just make the multiple things like the jumping jacks project that yesterday the Ayush sir has explained you like how the person is just jumping and it is counting how many jumping jacks it has done in the same way you can also find out the yoga poses right you are doing the yoga and how many times you have formed these things right yeah so I hope that you have understand the pose detections right I think Harsha Sharma has given an assessor there are 62 landmarks no Harsha Sharma in the face recognizations there are predefined 68 landmarks yeah okay fine so yeah so I hope you have enjoyed the part of it yes I hope you have enjoyed how we can make the pose detection based different projects and all right you have learned a wonderful things over here how we can make a different projects in machine learning right how we can make how we can just use the pose detection related things right all these things we have learned in a very well manner so now in the same way you can just try to make a different projects into your picture blocks application or in your dot like on the web version you can make a multiple projects like in the way we have just make the face mask detector you can try with the animal breed detector right you can try with that even you can try with the waste detector like whether the waste is a biodegradable waste or non biodegradable waste right so you can perform multiple activities multiple activities but performing the activity before performing the activities you will really need to know about how the things works in the picture blocks right so just explore the things learn the things and try to make the wonderful projects and you are going to learn a lot of things yes so now if anyone is having any doubt questions so you can put in the chat I am here to tell you it is not coming to upload the image model so like what you need to do go on the create a model from there it will redirect you to the teachable machine and on the teachable machine there is a three option image model prose model and audio model so you can choose the image model if you are having any doubt ask yes we are going to send you all the links again don't worry about it all the links replay link work link all the links you will be going to get don't worry about it yes yeah so thank you Shalini let me just give you a little understanding to all the students about the competition as well so as you already know that we are also having a competition going on so let me just share my screen and show you little about competition although we have discussed about this on both the both the days so students you already know that you have this is the last day for this camp I should say this celebration of scratch week with pictoblocks after this now we want you all to come up with your own project so you have been attending this session for all three days it's really wonderful we thank you for that and also now we are going to share you the link on the chat box and also you will be receiving it on your emails so you have to submit your project video which project video link from youtube basically okay and I just share you the link just give me a moment before sharing the link for the submission let me again remind you of the wonderful prizes you will be getting so you will be getting a prize a quirky innovators kid and the certificate so all the students who have attended will be receiving the certificates only once they submit their project video link which has been uploaded to the youtube to the form we will be sharing to you so let me now stop the share here and just share you the link and just hand over the stage to Jennifer to just let us know more about the quality hello everyone I hope I am audible to you yes okay first of all thank you so much Ayush and thank you so much Mr. Charlene for such a wonderful explanation I hope everyone would have understood that what machine learning is how to train the machine by performing the small activity also you might have got a decent idea from the webinar series that what is programming its benefits along understanding what AI is and the different techniques of its use the replay link of the entire learning coding from scratch will be sent in the chat box so that in case you missed out on the previous webinars or if you want to go through them you can go through them again as you must be aware by now that we don't want to preach we don't want just to preach theory to you we want you to apply it by making projects of your own and submit them to the competition and as Mr. Ayush mentioned the winners will be receiving and exciting AI and robotics kit quarky so let me play a small video for you this is rolled into one hi quarky it's an AI and robotics learning companion for the next generation quarky takes children on an exciting journey to explore advanced concepts on AI and robotics with an easy engaging and hands on approach it is packed with interactive online courses and engaging live sessions that help children to learn code and make their own real world projects in other words one quarky endless possibilities quarky is portable carry quarky around and control it using your laptop tablet or even a smartphone it's easy to program whether you're a coding newbie using blocks or an expert using python comes with a plug and play interface making it hassle free and with some of the coolest latest tech so imagine a self-driving car which identifies sweet signs and pedestrians or a faithful robot dog that follows commands with the help of machine learning and object detection it's evident that quarky is for everyone who wants to be ahead of the curve ready to explore the universe of AI and robotics with me I hope that video was very interesting for you a few minutes to talk about it quarky quarky is a reprogrammable AI robotic kit that allows you to bring your code to life it's a pocket size board to help you make a variety of AI and machine learning projects from self-driving car to waste collectors from weather monitoring system to expression replicators and what not and you can make these wonders from the comfort of your home we will be launching quarky globally on Kickstarter about a month from now and we want you to reap the maximum benefits from the same and to ensure that we are inviting you for joining the exclusive subscribers list you can join the same by registering by registering yourself at the link given in the chat box the learn code with this learn coding from scratch webinar series ends here but the scratch week does not you don't have time till the 20th May to submit your projects and stand a chance to win the AI robotic kit I'd like to thank all of you for joining and would also thank our presenters Mr. Ayush and Mr. Shering for sharing their knowledge thank you everyone stay safe thank you everyone thanks Jennifer for providing such a good information for the student teachers and parents the students as we mentioned please don't forget to submit your projects on the link given below I will be like stopping your chat and sharing all the links again so that you can keep it with you just paste some of the links the links again to download the picture blocks is here in your laptop and the pc you can just go to google play store and download the picture blocks in your android phones with this link to explore the web version of the picture blocks please do not forget to visit our website that is pictureblocks.ai and of course the most important please submit your forms please submit your youtube video links on the form so hope you have all enjoyed the session and we will be contacting you soon through the email also for these links and you will be receiving the emails and you will be receiving the links on your email do not forget to go through with them so I think Jennifer has posted one more link for you all students do not forget to register the quirky subscriber link here on the chat box thank you all for joining on this session thanks again bye bye everyone everyone