 So very good afternoon to all the educators who are joining us. Welcome to the fourth session of artificial intelligence bootcamp for educators. We'll be shortly starting, we are just waiting other educators to join in. Hello, very good afternoon to all the educators who are joining us. We'll be shortly starting in few minutes. We are just waiting other educators to join in. Again, a very good afternoon to all the educators who have joined into the session four of artificial intelligence bootcamp for educators. Also, like I would like to tell you basically, so it was amazing to see your projects yesterday. And we are going to display some of the projects from all of you. I think till the time other educators are joining, I should give them a short recap. Yeah. I'll just share my screen back. Hello, am I audible? Yes, I am if you are audible. I'll just share my screen. So yesterday we started with session three where we learned about the machine learning. But we also saw some of your wonderful projects. Okay. And then we started with the machine learning. So in machine learning, we understood that machine learning is none other than a subset of artificial intelligence which basically focus on the experience learning. Then we also learned about that how exactly it is related to human learning process. So the basic human learning process were divided into four steps. So we understood that with an example yesterday and then we related it with the real machine learning. So we were able to do that because we are like we are able to learn the things because we have certain senses, right? But machine don't have those senses. So that's why when we talk about machine learning. So basically there are three major tips which happens here is like we give the input to the machines. Then there is an algorithm working in behind which do the calculation parts. And after that comes up the output. Okay. We also saw the machine learning lifecycle. And we also did a hands on hands on hands on activity on this. So here you can see there were four steps where we just first gathered the data. Then we trained the data into the teachable machine. Then we calculated that like we just tested the data on the preview section. Once we were satisfied, we exported that model, right? And then using a link, we deployed that model into the crypto blog and we made a project of mass detector. After that we also discussed the application so where this can be very much used, right? So it can be very much used into the animal classifier, fraud detection, post detection. So some of the educators I saw in the feedback that they were asking what exactly is the post detection. So post detection is basically a technique where machine is able to detect the post, right? So here you have to think that I was giving an image. That's why I was able to like I was able to classify with the different images. If I want to make my machine learn about the poses, then in that case I have to train my model that is teachable machine model with the pose. And that is very much very easy. And then we want you to explore that into teachable machine and bring the modeling to the crypto blogs and complete the project. We also saw that it can be used in the healthcare and it can be used for the purpose of detecting pneumonia, okay? With the X-ray of the patient, it is used in recognitions, also recommendation system. You might have been seeing some of the things in YouTube or in Netflix that the things what you love to see automatically come in front of the screen, right? So these things happens due to the machine learning, okay? And it is called as a recommendation system where the machine automatically recommends you certain type of movies, the books, or the videos on the basis of which type of videos you are watching. So after that, we also made a model into the teachable machine and we like worked on the crypto blog. So I'll just stop my share here. And I would again like to welcome all the educators from around the world for joining us on the fourth day of the artificial intelligence bootcamp for educators. I'll just share my screen back. So we will be shortly starting all educators. We are just waiting for other educators to join us. Yes, I think we can start. Okay, so I would just again welcome all the educators from around the world to the fourth session of artificial intelligence bootcamp for educators. So today basically you already know that we are going to learn about natural language processing some how exactly we are processing the language on which we are talking and machine is able to understand that and then it is able to perform certain activity. So I will just request Mr. Pankaj to please start and let us know about the today's session. Yeah, so welcome. I will just do the screen share. So welcome back to the to the fourth session of artificial intelligence for educator bootcamp. And we are very excited to go into the fourth session of the artificial intelligence bootcamp. And we are very great, like we are very, we are feeling very great to see the response that is coming from the participants as well. So we are we are very excited and let's go ahead and see what's inside the fourth session of the bootcamp. So this is just a reminder all of you, most of you have already done this particular bit. So if you haven't joined the telegram group, you can join the telegram group from like using this link. We are keeping update adding all the updates here there and we are also solving the doubts as well. And if you haven't installed or downloaded and installed pictoblocks, you can do that. That thing as well because it's very important that you have pictoblocks installed in your devices so that you are able to complete the assignments and also practice the concept that we are like teaching to you as well. And I am like last last session also we have shared with you that we have a pictoblocks community on Facebook. So we want all of you to join in that particular community. The link of the community would be in the chat. Ayush is sharing that the link as well. So we would request all all the educators to join there. We keep on update like we have all the updates coming on in long term basis like not just for this bootcamp but for our future programs as well. And there are a lot of educators who and students as well who come up with a very good projects. They share their projects there and we discuss about the different opportunities, the possibilities about the projects as well in the community. So I would request all the educators to join us on the pictoblocks community as well because that is something which we are very much focusing on in growing the community and getting this culture ahead as well. So we are very focused towards artificial intelligence and robotics. So if you are like going to work on those particular things with students, I think you should join the pictoblocks community group. So coming to the project that you have submitted from like a lot of educators have submitted the project. And these are the projects some of the projects submitted by the educators. And as you can see all like most of them are very excited and happy as well. So very good. Like like I congratulate all the educators who have submitted the assignments. And it's great to see the response that are coming here as well. Some educators have put like gone one steps ahead and they have created their own custom models as well in pictoblocks and then created projects around it. So like as you can see there are models which classify the grapes like the fruits. There are models which classify whether you are like wearing the glasses or not. So there is a lot of things that you can do with pictoblocks with machine learning. And this is the thing that we've also wanted most of you to try as well. And I just want to show some of the projects that were created by the educators. Like it's a very innovative project that Ivansa has created with machine learning, getting like analyzing the posture and then do like classifying the posture and communicating with like for people who cannot speak as well. So it's a very good application that Ivansa has created. Coming to next project. Hello from Croatia. I didn't introduce myself very well in previous videos. So I'm going to tell you that my name is Kristina. I come from Croatia and I'm a school teacher in primary school. I teach students computer science and coding from the age of six to the age of 14. I'm also a mother of a six year boy who adores toys and cars. So I made this project extra project for today assignment where I use machine learning, which will help me to recognize cars. So here how it works. Hello, I'm your personal assistant and I will help you recognize your child cars because I know it is hard for you to test me. So let's go. I'm going to just do this. Okay, great. Here it comes. So as you have already seen the projects shared by the educators and you the people who have joined the telegram group, they're also like educators are putting up the new the project that they are creating as well. So it's like machine learning is a very versatile like versatile and you can create a lot of applications using machine learning. And these are the few application that the custom application that was created by the educators and the. So imagine if you give this tool to students, how would they respond to it? And like thinking about is it is also very excited us that okay, so if students get this tool, they would be doing a lot more. They would get more creative and they would also experiment a lot using these tools as well. So yeah, just let me set the screen here. So coming to the next part, so all of you already know what are the things that are on this like at state. So we have like we have been like giving you assignments and till now we have given you three assignments to complete. And the fourth assignment will be to giving you today and on completion of all the four assignments successfully, we will be giving you the certificate and the batch of Stempedia Certified Artificial Intelligence Educator as well. And the people who are submitting the submitting the projects or the assignments on time, which is like before 2nd of July till for PMIST like it will be one hour from now tomorrow. So at this particular time, they will be considered for the certification and for the batches. And top seven performers, we will be going over all the videos. Top seven performers would also get the opportunity to have the artificial intelligence kit with them as well. So we will be giving seven educators the AI and robotics kit to them on which they can do more than 30 activities and they can explore the world of AI with robotics and those parts. So it's very important that you submit your assignments on time and all the links you can find in the chat. And we will be also giving the fourth assignment today. So we will be telling you about that particular part as well. And we expect that you and we are also and we don't have any issue in giving you the certificates as well. So please submit all four assignments by tomorrow. So if you haven't submitted the previous assignments, then you can also submit it by this particular time, which is 2nd July 2021 till 4pm IST. The link of the assignments, you will get into the mail as well and you can also get it from the chat. And people who have submitted the wrong link into their assignments, they have to submit the assignments again. We will be notifying the people who have submitted the wrong link as well so that you get a fair chance to resubmit the assignment. And we will be looking forward to all the submissions as well. Now coming to the part now, which is the natural language processing. So today we will be talking about natural language processing and natural language processing is something which we are actually using a lot. If you see Alexa, then Alexa uses natural language processing at the back end. So what it does is if I just tell you about natural language processing, so I am speaking something. And you are able to understand what I am speaking and able to tell what to do just a second. Sorry for the delay. So currently I am speaking with you. I am telling you something and then you are able to understand the language that I am speaking with which is English. You are able to make sense what I am speaking of. And this particular part is something which comes under the domain of natural language processing. How we can make computer understand the context from the speech and put that context into actions as well. So let us go over and see it in more details. So first I will talk about how human language works. So we as a human, if you have to learn a new language, how would you do? So you start off with learning about each and every word. You start with characters and then you start with the word and then you put down the words together in order to make sentences. And now what you try is with the meaning of each and every word you put down the sentences and you try to make sense out of it. So if I say a sentence then what you do is you break down the sentence into the word and you make out the sense out of it. So for us because we are very intelligent, our brain is very intelligent, is well programmed or well trained to do it. We have that thing in us, the human intelligence. We are able to understand it very easily. But for computers it is very hard to do. Computer understand only zeroes and ones. It is very hard for a computer to understand beyond a certain limit. So what we have to do is we have to simplify all the things and we have to simplify all the long sentences into a computer understandable form as well. So this also what we do is we try to break down the natural language processing similar to how a human would process a sentence. So if I say a sentence, my name is Pankaj or maybe in order to simplify it, there can be a positive intent like today is my birthday. Suppose something is like this come or today I am sad. So these are the two sentences which are very drastic to each other. So which is like one is exciting, one is like sad and those kind of things. So how does a computer make the sense out of it? So now if you see today is my birthday is something which involves a word birthday which is associated with a positive intent. What we try to do is we teach the computer, we give the computer hundreds of sentences and we train the computer that if these sentences are coming in this particular, like if these words are coming in some particular sequence, then they are towards positive intent. If some particular words are coming in the bad intent, negative intent type, then those words are associated with negative intent. So here birthday is something which is a positive intent. So what the computer will do is computer will search for the words which are towards the positive intent. Computer will search towards the word which are for negative intent and what it will do is it will analyze the whole sentence and it will try to find out what are the positive intent, how many are the positive intent based words and how many are negative. And that's how computer what it does is using the algorithm, it figures out the positive and the negative intent of the sentence and it gives us the final result. So if I say today is my birthday, it would give me a positive intent. And if I say I am sad, then it will give me negative intent. So this is one of the process that you can do for associating with a positive or a negative intent. But when you try to extend this particular example, what you can do is you are basically categorizing some words that are coming in some sequence into categories. So let me give you an example. If you see Alexa, if I want my Alexa to turn off or to turn on the light, what you will do? You would be able to, what you would do is you would say Alexa, Alexa, I am feeling sleepy, please turn off the light. Okay, so this is the sentence that you have given to Alexa. But Alexa haven't like gone through and understand all the concept, like does not understand the whole word. So what it does it using the previous past experience, it tries to find out if the sentence is more towards light on or it's more towards light off. Okay, and with that particular decision making or classification, we try to do the computer try to identify whether it has to make the lights on or off. And this is something which comes as a part of natural language processing. Okay, so basically in natural language processing, we try to classify the sentences into some classes, which can be like lights on, lights off, fan on, fan off. It can be other things as well, like positive intent, negative intent, all those things comes under the natural language processing. So here the AI is actually working on the computer, like into the language part, which is like how we speak and those particular part. So what I will do is, okay, so in the natural, so in picture blocks, we have an extension, which is natural language processing. But this natural language processing is only available in the desktop software. Okay, it's not available in the web application and it's not available in the other Android or iOS application right now. So for this particular session, we will be doing the activity of speech recognition. Okay, so let me tell you about speech recognition first and then we will go over the activity and we will make a project on using speech recognition. And then I will show you the things about some demos of natural language processing as well. So in order to start, I will share my picture blocks screen first. I hope you are able to see this view. So this picture blocks project is something like I have created the whole background of it. So you have all the costumes and all those things created already. So today what we will be doing is you will be starting the class, you will be starting the session from this particular part where you have all the sprites available to you. And then what you are doing is you are, what we will be doing is we will be adding some like speech recognition part into the project. We will be focusing on the concept in this. So we will avoid all the part of like getting the costume and all those things. So we can just directly jump on to the part which is important, which is how the speech recognition works in picture blocks. So that would be the part which we would be focusing more on. So I have put down the project link into the chat box. So you can also download the project from the chat box. So the first step that you have to do is you have to open the file that I have shared with you on to the chat. So can you just share the files again into the chat? Yeah. I hope my screen is pretty visible to you. Yeah, it's visible. Or I can zoom in more. I will just check it. Okay, I have a little bit and maybe over zoomed it's okay. Yeah. Okay, so now you can see the part here. I will just make the stage a little bit smaller so that we are able to see the core. Okay, now let's go over and check what all things are here. So what we have is we have all the states set. So the scenario here is here we have a stage where there is a girl with name Abby and Abby is in home and she is actually talking to a personal assistant like something like Alexa talking to Alexa. She would talk to Alexa and what we have to do is we have to give her the instruction, whether to turn on the fan or to turn off the fan and whether to turn on the light or not. So there is a lamp here and then there is a fan also here. So the script of the fan and the lamp is already written. So we don't have to actually write the script of the fan and the lamp. We will be using the pre-built script that we have created already. So I use keep the project link pinned into the chat box so that everyone can get the project link as well. So this is the starting point that we are having. Now let's come to the objective. So our objective here would be to so Abby would ask us whether to turn on the light or not. Okay, so Abby would ask us what do we what do you want to do? Okay, and then using speech recognition, we will be telling to Abby the intent that we have. Okay, so I can say fans on or I can say lights on. Okay, so if these kind of intents are I am giving to Abby, Abby will turn on the fan or turn on the light or turn off the fan or turn off the light according to whatever instruction that I am giving. So we are making a voice based assistant system where we are giving the instruction using our voice. Okay, and this is something which we will be doing in this particular project. So before going to the speech recognition part, I would just tell you about the concept of speech recognition. So speech recognition is something which all of you have must have seen the scene. Whenever you install your windows, you get Cortana, you have Alexa, you have Siri, a lot of speech recognition based applications are there in the thing. And yeah, so lamp sprite is also there. So you can get the lamp sprite there. So so a lot of speech recognition recognition based devices are there. So what they do is they will be getting they will be getting a lot of what they do. They listen to you first, then they go over the record. They have an algorithm in order to break the whole sentence into different part, get this label like ABCD and then join the things to get to make sense of the sensible words and then the sensible sentence. So it's a very, very, very like good for how to create like speech to text. So it's a process where the computer first find the characters, then it form the words and then it form the sentences. So these speech recognition based devices are a lot into the market and we will be creating a similar device in the pictor blocks today. So in order to get this to start off the application, the first thing that you have to do is you have to go to extension add extension. You can see the add extension button here. Click on add extension and get the artificial intelligence extension from the library. So you can see the artificial intelligence extension here. You just have to click the artificial intelligence extension and it will come to that. So the speech recognition is actually available in artificial intelligence. So you have to add the artificial intelligence extension in order to go see this thing. So when you get the artificial intelligence extension, you would be able to see the blocks of artificial. You have already used it in image features and those part. So I won't go to that part again and again. But when you see the blocks, then you will be finding a block which says recognize speech for two seconds in English it says. So it's basically using the speech, like recognize the speech. It first record the whole speech and then recognize, convert the speech into text. So this thing, like we perform this particular task to get the text from the speech. So what you can do is there are two parameters here. So as you can see here, the parameter that we have is the length of the recording that we want to do. So it can be two second, five second, 10 seconds. I would just put it as a five second recording. And then you can also select the language in which you want to convert. So there are a lot of languages in which you can convert the speech. So you have like Arabic, English, Hindi, all the language, Gujarati, a lot of languages are also there. So what I can do is I can just click on the English and I would just try this block running this block and see what the result is. So when I click on this, hello, my name is Pankaj. Hello, yeah. Now the recognition window is coming and you can see that my video would also come. So it took some time for the first time. So I would just, yeah. And you can see it's getting this particular part as well. So where the speech is being recorded. So I will just try it again. Hello, my name is Pankaj. So what it has done is it's recorded. It has recognized the speech. And whenever it stops glowing, it has recognized the speech. And now we can move on to the part where we can use the text from the speech. Okay, so as you can see here, we have the block which says, like we have recognized the thing. And if you just go in the extension, scroll down, then you will be finding speech recognition and then the block is speech recognition result. So this particular block actually gives us the result of the speech recognition that we are doing. And if I just click on it. Okay, so it has said, hello, my name is Andre. I can just try it once again. Hello, my name is Pankaj. So Pankaj's name is not very common in English. So that's why it has tried to make a word which is available in English. So that's why it's saying this. So it's doing the speech very well. You are able to get the speech converted to text. So this was the first part that we wanted to do. So we wanted to have a speech recognition created for us, which can just recognize the speech and then tell us what we have said. Okay, so I will just test it out with one more thing which would be, let's say I want to turn on the light. Lights on. It says lights on. Okay, I can also try lights off. It says lights off. I can also do a lot of things which can be like converting to Hindi. Hello, my name is Pankaj. So it would convert in Hindi. And so it works very well in different languages as well. You can try to do it. And if you are, yeah, so you can do it and try to do it in different languages as well. So we will stick to English. And what we will do is we will start off with the program here. So I will just close the recognition window. And as you can see, what I have is I have the code written for, so let's go over and see what are the codes that is written for lamp and fan. Okay, Ayush, you can just put the link of the file again so that everyone has the file and they are able to do the activity. So when you go over the code, you can see that for the code, we have a lamp. So for the lamp, what we have is we have, the code is very simple. When the green flag is pressed, the lamp goes to a certain position. It sets to a certain size. And then it repeats this particular code, which says, if the state of lamp is on, then turn on the lamp. If the state of lamp is off, then like off, then we have to turn it off. So this is something which is like a basic, what I would say, basic convert, like turn on the light or turn off the light. So it's very basic. So what we have is we have two variables created, so fan state and lamp state. So the lamp state actually tells us the state of lamp and the fan state tells us the state of fan. There are other methods to execute it as well, but we will be sticking to this particular, this method of using variables. Okay, so we will turn on the lamp or lamp on or off using this script. Then the next script is for the fan. Okay, so the fan would actually rotate itself. So it would do something like this. So if I just go to forever and just get this thing here, it will do something like this. So that we are getting the visual of, okay, the fan is actually rotating. So this is something, a small animation that we have created. We have a different costumes for the fan, which is allowing us to do this particular part. Okay, so now as you can see here, we have to set the position of the fan. We have to set the size of the fan. And now when all the things are done in the forever, we would be doing, if the state is on, then change the costume. So we have an illusion of fan turning. And then if the state is off, then what you can do is you can just stick to one costume so that the fan is actually off. Okay, so this is something which we will be doing for the fan and light. So what we have done is we have the switches for fan and the light. But we don't have the person switching on the thing, but we will be doing all those things with a voice activation. So in order to start that part, I would just do one thing, which is if you go to Abbey, we have a small script written for the Abbey as well. And Abbey script is basically setting up to the position. So the script is very, very straightforward. It sets the Abbey position to somewhere in the stage and then set the size of Abbey as well. So we can increase the size of Abbey and do all those particular parts. So this is just for making the scene very pretty. Okay, so the file that you have got is the basic setup for the project that we will be creating today. Okay, now the next part that we have to go over is, what we can do is we can, whenever we are starting the program, we will first do the part of recognizing the speech for the particular time. So what I will do is I will get the recognition window here and I will just click or get this particular block here. Okay, so what is happening when the green flag is clicked, recognize the speech for five seconds. So the speech recognition would have happened and what you are getting is you will be getting the result. Now, what you have to do is based on the speech recognition result, we have to make decisions. And these are the decisions that we have to, that will act as a person switching on or off then and switching on or off the lights. So what we will do is we will start off with the first part, which will be switching on or off the lights. Okay, so I would just go to, so I think it would be very easy for us to do because in order to turn on or turn off the light, we will have these things. Just a second, let me zoom a little bit. Is my screen visible very clearly? Yeah, not visible. Yeah, okay, okay. I guess you can share the file link in the chat as well. And if you are not able to go like make the project right now, you can try it afterwards as well. So that won't be an issue. And just to take a quick feedback, have you all understood the concept discussed till now? If you are facing any problems in the extension of those particular part, you can ask in the chat and Ayush will be helping you out in the chat as well. Yeah, I think most of you are understanding. If you are not able to understand some particular part, you can ask into the chat or you can go over and see the recording as well. So you will get the recording onto the YouTube channel. Yeah, so coming to the next part, now we have to make decision based on the speech recognition. Okay, and here because we are using text as a text, we will be using some operators from here. Okay, so what I will do is I will just get one operator, which is contains. Okay, currently we are not implementing NLP into the project. I will give you demos of how you can implement NLP as well, natural language processing into the project as well in order to make it more easy for people to give a various command. Currently, I am focusing on speech, once I get the speech recognition result, how I can turn on the lights or turn off the lights. So the first objective is to control the light. So what I can do is I can just go here and I can see that, okay, if the speech recognition results contain off on. Okay, so let me just try. So lights on. So my speech recognition result is actually lights on, light on, and does that contain on? Yeah, it's contain on. So if this particular thing is happening, so what I will do is I will just go to control, and I will get a block, which is this block, which is if then, and if the light, if it contains on, I will just, what I will do, I will go to variable and just change the lamp state. So if I can just set the lamp state, I will get this block here, I will just get the lamp state to on. So if you remember, if in the lamp we had, if the lamp state is on, then the light would turn on. So here also what I have to do, if the speech recognition result contains on, then set the lamp to on. So I will do this particular thing, and then I will also create like what I can do is I can just duplicate this particular part. And I can see if the speech recognition result contains off, I will set the lamp state to off. Okay, so this is a very, very easy if statement like conditional statement. So if the speech recognition result is on, then contains on, then I will turn on the lamp. If it's off, then I will turn on the lamp, turn off the lamp. Okay, what I will do is I will check the thing, check the part that we have implemented here as well. So let's start work with lights on. Now you have seen that the light turned on and lights off. Okay, for lights off, what did I get? Okay, it got light on. Lights off. Yeah, so the lights off came and then light turned off as well. So this is a very, very simple conditional statement that we are doing. And there is one particular thing that you would see here is basically what we are doing is we are hard coding it. Okay, once the speech recognition result has come. But if you are using natural language processing extension and create a classifier of your own, then you don't have to do it. You can actually create class and then do the natural language processing based classifier. Use the classifier to see whether to turn on or turn off the light as well. This is what I would say a part where we are focusing more on the speech recognition rather than the natural language processing. I would show some demos about that as well. Yeah, so as you can see here, we have implemented the thing for the light. But if I want to do it for the lamp as well, for the fan as well, then you would think that we would require some modifications here. The first modification is if I have to turn on or off the fan, then I would say again on or off. Okay, so it won't work in this particular manner. Like on is on for both fan and lamp. Okay, so what I can do is I can add a conditional statement which can actually be first seeing whether the speech is actually meant for the lamp or it is meant for the fan. So first we can put the focus there. So what I can do is on top of the both conditional statements, I can put this particular statement like if. So first I will check whether whether the speech recognition contains light. Okay. And if the speech recognition contains light, then I will change the lamp state. Okay. So as you can see here, what I'm doing is something for the something contains for the lamp for the lamp for the light actually. So as you can see here, what we have done first we will check whether the speech recognition result contain light. If it contains light, then we will change the state of light. Otherwise we will not change the state of light. Okay. So if you if you're finding difficulty on getting this operator block, you can just click on the operators and get the contains block from here. It's basically checks whether the the first input has the second input or not. So it has it says true if it has if like I see me it says false. So so it's a very easy to understand block. So now what I will do is I will just test this program once so that I am very sure that I should be sure that the program would work. So I would just go here. I would just start the program lights on. Yeah, the lights are on lights off lights are off. Okay. So on. So we have to also specify whether we are turning on the lights or not or turning off the lights or more. So this is something that we have to specify. Now if you want to implement it for the fan. Okay. What would you do? You would basically have to create the whole program but specific for the fan. And there is only one part that you have to change like three parts where you have to change. So what I will do is I will go. I will just duplicate the whole. So I will duplicate this particular section. I want to duplicate the whole section is because the whole section would be so the previous section was created for the flight. The next section would be created for the fan. So once I create this like duplicate whole section and put it below the if statement. You don't have to put it inside but you have to put it below. So it's very important that you understand the concept of whether to put inside or the outside of the loops. So you have to put it outside because this is a completely different statement. Okay. So if the speech recognition result contains fan. This is first change. Then I would change the lab or the fan state to on and the fan state to off in these particular positions. Okay. So what we have what we have done here. So maybe I should just give me a second. I will try to get all the things in one space. Yeah. Yeah. I think some of you might be able to see this particular part. So it's basically what you have is your test is for light. If it's contained light or not, then you are tested for on off and then do it for fan as well. So let's see the result of the project that we have created fan on. Yeah. The fan has started turning turn off the fan. Oh, sorry. Turn off the fan. Okay. So turn off the fan. Okay. So in this case, we haven't got the result that we wanted. So we wanted the fan to be off, but it's still on. So there is a there is something which we have to understand. So if we are finding out difficulty in getting the right result, we have to focus on the debugging part as well. It's very important in terms of getting yourself prepped for the for the in the project making process. You have to debug a lot of things as well. So what I will do is I will just make the screen bigger just to show you how to find the problems of problem here. So I will just go to artificial intelligence. And if you see the speech recognition block has a check mark here. So when I click on the check mark, it actually say gives me the what I'm getting. So it's written that turn off the time. Actually, it hasn't haven't turned off the like recognize the word fan properly. So maybe I have to put like, say the word properly here. So what I will do, I will just go again here and I will just start the whole script. Turn off. Sorry. So I have to start it again because I was already doing and then the recognition has happened already. So turn off the fan. Turn off the fan. So let me try it again. Turn off the fan. So let me fan off. My accent is not working here. Okay, I actually like gone to the UK part. So I will just try it again. Turn off the fan. So it turned off the fan. So my accent is not that towards the UK English but more towards the US English. So that's why it was not recognizing the accent that I was having. So yeah. So this is something which like you would be able to try it out. You can like this is a small project that we created where we are shown. We have shown to you the power of speech recognition and then create something like lights on or lights off or fan on or fan off type of things. So if you are like this is something which you can do it in more like languages as well. So you can make speech recognition in other languages as well. Now in just in order to create add more things into it. You can also make a Abby say something to you as well. So I would like you to show you one more block. When I click on add extension I go to the extensions then you would have one extension which is speech to text. So this actually is something which is opposite of speech recognition. Okay, so the opposite of speech recognition would be would be something like this. So it would speak hello. Are you able to like get the voice from the computer? Hello. Yeah, yeah we are able to hear that. Yeah, so it says hello. So what you can do is you can make Abby say something to you as well. So hello. What can I do for you? You can say and then you start the speech recognition. So just to just to show you how it works. Hello, what can I do for you? Turn on the fan. Turn on the fan. Now the fan has started. So this is the project that we wanted to we wanted all of you to make. And yeah, now coming to the natural language language processing part I would show you some of the projects that you the extension of the project with a video. So before that I would just ask questions. Have you understood the concept discussed? No, it's great to it's great to see a lot of response coming as yes. So, so yeah, great, great. If you have any issues coming on, like if you are seeing null, then please check your ones like you can check your internet maybe once and maybe you can see your credit as well if it's not working for you. If you if it's coming null, please check the pictoblocks credit as well and if you require more credits then you can directly like send your request to support at the rate does tamperia.com we will be providing you more credits for doing and practicing other projects as well. So if you are not able to get credit just mail it to support at the rate does tamperia we will give you the redemption code the new redemption codes as well so it's based on the requirement that we are giving. Okay, so now let's go ahead and just see a video of the natural language processing and yeah. Hello everyone, I hope you all are doing well. Artificial intelligence is gaining popularity day by day and the most fascinating is natural language processing using which robots can understand the languages spoken by humans. Not only understand but also feel the emotions behind it. Isn't it amazing? In this video, we are going to try our hands on on the latest NLP extension of pictoblocks by making virtual doctor. You just need to tell your symptoms and he will tell you what caught you and what you should do next. Along with the NLP extension, there are a lot of new features added to the latest version of pictoblocks. If you haven't downloaded it yet, you can download it from the link given in the description box below or the card above. Let's begin. Open pictoblocks. First of all, let's add the doctor and the hospital sprite and set up the stage. You can download these sprites from the link given in the description box. Once downloaded, you can place them as per your choice or like the type we have placed here. Next, add the NLP extension. NLP stands for natural language processing. This is the area of AI concerned with the interaction between computers and humans in natural language. The aim of this extension is to help computers understand language as well as we do. Now, we need to train pictoblocks to analyze the text and perform the corresponding actions. For training, we will add data of symptoms for three diseases, that is COVID-19, malaria and diabetes, one by one. Go to the MyBlocks palette and create a block named data for COVID-19. To add data to the text classifier, we need the NLP extension. Take an ad class block and place it below the ad block. In place of text, write the first symptom, dry cough and write COVID-19 as class. Let's add more symptoms to COVID-19 class. You can add as many as you may find. Only thing you need to do is just duplicate the block and write the symptoms you want to add. Now, similarly, add data for the other two diseases. We'll add three symptoms for each. Now we need to train our text classifier. Create a block named training. Take the training ad block and add a reset text classifier block below it. Add the data for COVID-19, data for malaria and data for diabetes tag blocks one after the other. Add a train text classifier block. Next, let's begin to write the script to make the virtual doctors say the name of the disease we have after we tell him our symptoms. Add a when flag clicked hat block. Drop the train data tag block below it. Let's make the doctor greet us. Add a save block. Write, hello, welcome to the virtual hospital. Next, we'll want the doctors to ask the symptoms. Thus, from the sensing palette, add an ask and wait block. Write, how are you feeling today? Please tell me your symptoms. Here is when we will try the symptoms and the doctor will tell us the remedies for the same. Place enough block into the scripting area. Add an equal to block from the operator's palette. Inside the first space of the equal to block, drop a get class of block. In the second space, write COVID-19. Inside the get class of block, drop an answer block from the sensing palette. This will check the symptoms you enter and if they match with COVID class, the doctor will say out the disease as COVID-19. Into the if arm, place a C block and write, you are infected with COVID-19. Now that we know the disease, the doctor will say the remedy for the same. Place a C block from the looks palette and write remedies into it. If you need to write more remedies, you can use another C block too. Similarly, we will do malaria and diabetes. Duplicate the if block and make the changes respectively. With this, our script is complete. Press the green flag to start the script. Let's test and write the symptom as, I am feeling breathless and have been experiencing fever. Yes, and got detected correctly to escalate the project. Instead of typing your symptoms, you can use speech recognition blocks to recognize your voice and then diagnosis. So this is all from my side. Now it's your turn to make this project. But don't forget to share this project with the picture blocks community on Facebook, which you can join from the link given in the description box below. If you like this project, give it a thumbs up and subscribe to. Yeah, so the video that you find found was there to show you the application of NLP. We couldn't cover all the, because of some limitation, not all of you would be able to do the NLP project. So we have covered the second part of it, which was the speech recognition. But we have videos on YouTube for the NLP as well. You can go over there and practice those parts as well. And as you saw in the video, what we did is we trained the natural language processing classifier or the text classifier on different categories. Same thing, you can do it here. You can train the classifier to do a lot of things. And then whenever you are saying something, it can make the decisions using the classifier as well. So I would go back to the slide and just move ahead in the session. So yeah, so you saw the project of a virtual doctor using NLP. So like you just have to train the classifier and then when you train the classifier and put the like put the classifier into the test, it will show you the result of the natural language processing. Now there are a lot of application of natural language processing that comes in because natural language processing allows us to give sentiments. So you can do a lot of surveys. You can have a lot of chatbots created on that. You can use NLP to shortlist the candidates for hiring and recruitment. You can do it for auto search and auto correct correcting the statements as well. So email filtering is something which is very popular. So all of you know like there are email filters who put the emails into spam or it puts into the inbox. So these filters actually identify and go for words and identify the words and the sequence of the words that are more spammy. So it identifies or it gets trained on it. And then once it finds those spammy mails, it puts into the spam and then puts email into the inbox if it's not the spam. So spam is one of the major application of natural language processing and using NLP and speech recognition, all Alexa Siri, all of them works on that particular part as you can see. There is one more application that you can do with the hardware. So what the application that we did for the like today on the lights on or lights off, you can do it with the hardware and using the speech recognition, natural language processing and the hardware interaction as well. So I would just show you the project of the same lights on and lights off happening on this hardware. I want to study. Sorry, the video was off. Yeah, so in this particular class if I hope you you got the you have heard the thing. So are you sure where you able to hear the the command that was that I was doing. Yeah, so we were able to see the complete working as well as your voice. Yeah, so as you can see here we have created a classifier which classify the things into lights on and lights off. And here's I specifically have not used the word on or off or lights or off. So actually I have trained it into it to get get classify the sentences into sentiments of whether I want to study or not, or whether my sentiments are towards like towards whether the light should be on or off. So I trained it in that particular manner. And when I speak when I use the natural language like an LP model to classify this particular thing, then I can use it for something like this. If I want to sleep, the light should be off. If I want to study the light should be on the same thing is happening here. You are actually not specifying the AI whether you have on you're not actually hard coding it. You are creating a AI which classify and we which classify the whole sentence into different segments, whether it's sure it should be light on or light off. So these are the projects that you can also make using the hardware and you can also do it in this sprite and those particular part user using natural language processing. So an LP and speech recognition go hands on hand and it is the part which is being used in a lot of application and student loves it. Okay, they try to they love to make a different kind of classifier. They love to play with the speech recognition. So this is also something which the students get more excited about when we teach them into the classes. Yeah, so now the next thing is the project hub. So the activity for today is complete and I would request all of you to yeah. So if your NLP is not opening, then like you might not be using your software. So the NLP is only available in Windows Mac and Linux software which you download and install it from the website. So it won't be available in the Android or in the web version. We are making the compatibility with the Android and the web version very soon. So I wish you can put down the link of the project hub in the chat as well. So we have a very good project repository where we have more and more types of projects related to the AI application. So you can see the projects such as virtual doctor using NLP, the face based attendance system. A lot of applications are already like we have created in our project hub. So if you want to like explore more projects and you want to explore more projects in terms of AI, then please visit our project hub and do and check out the different opportunities and the project that you can do. And you can also subscribe to our Stempedia YouTube channel. We have the project videos on our YouTube channel being published very regularly. We publish the AI based tutorials and projects very regularly. So if you want regular updates on the videos, then you can subscribe to our channel and also turn on the bell as well. So that you are being notified about that part. So let's come to the assignment. So today assignment is for you is to make the project on the speech recognition. So the project that we created for Fanon and Fanoff, that is actually the project that you have to create. You have to make the video of you telling the command and running the project and showing us the different variations of it. So like you can first thing is you have to make the project that we have created. And after that you can remix and add more things into the project and make the video of it and submit the project in the assignment. So the assignment link you would be getting into the chat and I will also show the link into the assignment link as well. So it's very important that you submit the assignment as a form of video. So the only format that we are taking is a video format so that we are able to see that you have actually done the assignment, not like just get the file and then put it onto the assignment. We want all the assignments to be genuine. So you have to record the assignment, the video of the assignment and upload the video either on YouTube or Google Drive or like we transfer or any other file sharing platform and share the file link with us. You don't have to upload the file but you have to share the file link with us. You can upload the file into YouTube and give it to us. So the file, you have to be particular about two things that we are able to access the file. So if you are uploading it into YouTube then do not make the video private. If you make the video private and submit the link as the assignment we won't be able to see it. So please submit an unlisted or a public video. And the second question, the second thing is if you are submitting the Google Drive link, then in the Google Drive link please make sure that you have a public view access of the file so that we are able to see the assignment and grade it as well. So can you just put the Pectoblogs file link as well? Yes, Ayush has put down the link. So the deadline of the assignment, so you can see the submission form and you will also get the submission form in the chat and the email as well. So the submission form is there and yeah, I will just go to the part which is very important. I think I missed that putting that slide there. Yeah, so the deadline of the assignment is actually 2nd of July. So we might like we have the deadline till 4pm IST. So you have to submit the assignment before that all the assignments before that we are able to grade all the assignments and give out the certificates and the batches as well. If you have any questions related to Pectoblogs credit, related to any issues or doubts, then just put the questions into the Telegram group or email us to supportadheritdustampedia.com We will be giving you the proper support in order to get your projects done before the time as well. So please make sure that you submit all the assignments. You have to submit all four assignments in order to get the certificate and the batches and be eligible for getting like artificial intelligence kit. You have to actually submit the video link only, the link of the video project only. Okay, so you have to create a video of the project working and then you have to submit the link of the video. You don't have to submit the link of the project. We want the link of the video created by you working on the project. And it has to be the same project has to be submitted. It won't be like you are submitting the third assignment project everywhere. Okay, we will be monitoring all the assignments. So please submit the assignments properly. And on completion of the assignment, we would be like giving the certificate and the artificial intelligence kits to the educators as well, top seven educators as well. So this is about the deadline of the assignment. Now coming to like you will get the emails about all the all the email like links on the email. So please make sure you have the assignment submitted on time based on the based on the response from this from like if you have any concerns related to the deadline and those things, you can just contact us at the contact us at education at the redusampedia.com and we can see the see the case if you have any problems related to the deadline. So we will take care about that part as well. So let me tell you about the session on the third of July that we will be having. So this would be the last session that we will be having on third of July. So third of July 3pm IST again, we will be having the last session. And this session is basically focused towards the learning opportunities and the course material and the like educators material that you can have for the students. So in the whole camp, we have only covered four concepts of artificial intelligence. But in that particular session, we will be talking about more concepts, more techniques and more application that you can do. And you would be able to like perform like you would be able to see the whole how you can implement the program on the third of July session as well. So it's a closing ceremony and we would like it would be a compulsory attendance for all of you who are attending the camp. And we would request you to come on to the on third of July in order to like get the updates about the who won the prizes who like and we will also be providing the certificate and the batches as well. We will be announcing the name of all the people who have got the assignments and the batches as well. So yeah, so so we will be doing that that part. So so this is something which we wanted to tell you we will be so there are two things that you have to do it now like there is actually one thing that you have to do it now. Go over the go over all the assignments, submit all the assignments by the deadline so that you are able to get the certificates as well. So it's very important that you are able to do all the certifications and those part. Yes. And if you have any questions you can join the telegram group. I would be taking the doubts now so there were a lot of doubts coming on to the project so we can take out some doubts. And one doubt that I'm getting is is any session tomorrow so tomorrow is not tomorrow there is no session. You have to work on the project. Tomorrow we have kept the tomorrow open for the project meeting part. There will be a lot of you who have to submit all the four projects so tomorrow will be a very busy day for all of you in order to submit the project and get the certificates as well. So we will be getting you all the things. Yeah. How can I get the recognition result. So I think the one question that I was seeing was how to get this text to speech extension. So one question was that so in order to get the text to speech extension you have to go to add extension. And then here you would be able to find the extension in the bottom so it's actually in the bottom of the thing where you have text to speech. So this text to speech is very you can find it here and click here and you are able to get the text to speech. You can also change the voice if you want you can change the voice to other people as well. So when you change it the voice will change. So like this is something that you can explore you can set the language as well. So if I just set language in the hello so so the accent is there but if I say namaste it would say namaste so so that is something that you can also explore. So you try exploring these text to speech block as well how to change the state of land. Okay so in order to change the state of land so how we added here is the lamb state was actually added into the variable. Okay so the variable is actually lamb state. So as you can see here on the variable we have the three two states which is fan state and lamb state. And you can see this lamb state right now is off. If I just go over it and do the thing it would stay on. Okay if you are getting null then there would be two reasons for that one reason would be either your internet is not correct. But if you are attending the video as well and doing the same thing in the device then you have not like this is this won't be a valid reason for you. The second reason that you would be having is that your crypto blocks credit are exhausted. So if your crypto blocks credit is exhausted you can just see it from here. You can hover over the icon here and you can see the crypto blocks credit you have currently. So if you want more credits we will be providing you more credits and we will be providing you credits on request. So you have to mail us to support at the rate just and be here to get more credits on that. Okay if you have like that we will be doing for you to complete the projects as well. So yeah I would also like to ask you to please just show that how we can see the recognition results. Yes yes yes one. So the recognition result is actually added in this particular block which is the speech recognition result. Okay you can put the speech recognition result on the stage as well by just clicking on this tick mark. So if you just see the tick mark and maybe I can put this thing in there. So you can see artificial intelligence speech recognition results. So it says whatever the value is inside it. So the email ID for getting more credit is support at the rate just and pdr.com. Ayush you can put the email ID. I think a lot of you would require that so we will be doing that. Okay how you can so yeah so you can get the recognition result from here. You can just see that I can just get the block here and you can find out when I click on it it gives me the results. So turn on the fan. The other thing that someone asked is how we can change the state of the fan. So to change the state of the fan we are actually setting the variable to some particular state. So we will set the fan state to on. Then the fan state would be on. If we set it to off then fan state would be off. You can see the fan state turning off when I click on there. Yeah. If it's giving you null then please check the picture blocks credit and email it to us and we will be providing you more credits in order to do the assignments and the activities. So we want all of you to be the very aggressive towards assignment. So we will be providing you support in terms of credits that would and also one more exciting thing that you would be getting. You will be educators who get the certificate as well. What they would get is they would get 25,000 credits on completion of all the assignments as well. So we will be providing each and every educator who completes the all the program, please all the assignment 25,000 credits as well when they complete all the assignments. So that is also something that you will be getting when you complete all the assignments and get the badge, get the certificate, get the credit. So there are a lot of things that you will be getting. So after that you can like you will be like you can practice more. You can develop the curriculum as well. So all those things we will be providing you and we are here to support you in terms of getting these things out as well. So if you require any help from us, any particular support related to the curriculum related to the course, we would be able to help you out. Also I would like to give some of the answers asked by the educators. So here some educators said like they are not able to access the NLP how they are going to submit the assignment. So today's assignment is basically NLP and speech recognition. So if you want to explore the NLP part, go and explore and make some assignment out of there. Although the educators who are using the smartphone, they can make their assignments using the speech recognition. So if you are able to make the things with the speech recognition just make it, there is no problem. And one more question was like, are we going to send the acknowledgement list? So yeah, we will be changing the acknowledgement list of the educators whose assignment have been like accepted. And if some there is some problems with some of the educators, so we are going to share you the acknowledgement list soon into the telegram group. Also like as mentioned by Mr. Pankaj that don't miss this opportunity on coming on the 3rd of July, the next session, because we have so many things on stake and we are going to have a wonderful ceremony on that day. So we want you to not miss this opportunity to take all the grab up goodies and everything. Also, we would request you all to share your experience video onto the telegram. So just share your experience with the boot camp, with the pictoblock, show your love and share us on the telegram group. Because we will be like combining all your experience and coming up with on the ceremony day. Okay, so please don't forget to share your personal videos of your personal experience videos with pictoblocks and the boot camp. Yeah, so there is one doubt how to switch off the light while it's on. It's taking time to switch. Does it happen like that or can you show us the execution button? So the speech recognition is something which would take some time. So if I just go here and show you, I would just select Abby. So as you can see here, it's like if it's on, if you want to turn it off, then this is the condition that would turn it off. So first there should be the light word and then there should be off word in the thing and then it would turn off the light. Yeah, so if you are not able to join the telegram, then don't worry. You can also look into the emails that we are sending or the update email and you can do that again. So if I just have to show the execution, I would just... Hello, what can I do for you? Turn on the light or it turned off the fan. Okay, just try and let me... Hello, what can I do for you? Turn on the lights. Turn on the lights. Yeah, I turned on the light. So maybe you are not able to see because of this variable. Yeah, the lights are on and if I just... What can I do for you? Turn off the lights. What can I do for you? Lights off. Yeah, the lights are off. If you want to get the fan speeds faster, I guess you have to maybe delete some sprites or reduce the wait time. So you already have the wait very slow. So like fan would be spinning at this particular speed only. So maybe you can try it out with yourself and see if there are any other results. Yeah, you have to click on the green flag and then only say because all the script are starting when the green flag is clicked. How can we get the lamp and the sprite? So if you haven't got the lamp and the fan sprite, it's because we have given you the Pictoblox file as a starting point. So you will be getting the... You can get the Pictoblox file as the starting point from the link Ayush is sending on the group. On the chat. Any other questions if you have? Just to give you a reminder, you have to submit all the four assignments by tomorrow. Yeah, so yes. So someone has suggested if we put the delay as 0.001. So that might help. It has increased a little bit. So tomorrow we don't have a class. So if required, we might do a doubt session class tomorrow. If some of the people, if we find that thing into the telegram group, we might do a doubt session on Zoom call or something like that. But till now we do not have any class for tomorrow. So if you think that you require more support in order to get some particular doubts clear, we can do that. So let us know if you have more doubts in the telegram group and we would be able to answer all the doubts there. So tomorrow is no class. You have to do the assignments and complete the assignments on time for getting the certifications and the prices as well. So I think we will be closing the session now and once again I would just put up the PPT. For you to access the. Yeah, sorry, sorry, sorry. Yeah. So, so if you want to contact us, you can like if you have any support queries, mail us to support at the rate the same media dot com. If you have anything related to some business inquiries you can do at bd at the rate the same media dot com. If you have any query related to the deadlines, education policy or those kind of things, then you can email us to education at the rate the same media. We will be taking out all the things we did all the emails very regularly and we would be answering all our doubts and queries there as well. So thank you everyone for attending the session. It was a great session for a great session where you learned about speech recognition natural language processing. And if you have any questions then put your questions in telegram group show your love, complete your assignments and also please make an experience video about the boot camp and share it on the telegram group so that we are we get motivated to do more and more of these kind of camps for future Stanford students in the future as well. So I'm very excited to see all of you in the closing ceremony that would be happening on third of like third of July at the same time at 3pm IST. So thank you everyone for attending the session and thank you very much. Thank you everyone for joining this boot camp and we would love to see you all back again on the third of July and we are just waiting to have your wonderful project videos and experience videos. We will be sharing you all the resources in your emails as well as on the telegram group. So thanks again. Thank you.