Building a No-Code Machine Learning Application (for Computer Vision) with Microsoft's Lobe

Key Takeaways

do you want to build an application using deep learning and computer vision but without having to code if you answered yes then this video is for you because today i'm going to be showing you a brand new machine learning tool that will allow you to do that and so without further ado we're starting right now okay so the application is called lobe and it is developed by microsoft and as you can see here there are several examples that they are providing so essentially this is a downloadable software so let's click on the download link and then you could type in your name and email and then you could register to download this okay so i will leave the download part to you and in the comments down below of this video please share with me what awesome applications are you developing today and so let me give you a very quick tour of this awesome application so today i'm not going to be doing any demonstration but i'm going to be just introducing you to this awesome application because this application has just been released a week ago and if i was to do a demonstration that would probably take me quite some time in order to do the filming and also to do the editing of the video and so in order to allow you to have access to quicker information i will be just giving you a quick round down of this awesome application so here on the website it says that you could train the application in order to recognize various tasks like for example if you're exercising you could count the number of repetition you could even teach it to detect emotion you could also teach it to detect color you could also create an application that would detect whether you're wearing a mask or not or also to identify plants and as you can see you could label them and also to identify the gesture of the numberings again this is counting the repetition as i've mentioned already so let's have a look further so here it's going to be quite easy as they're mentioning here and so this is a overview of some of the correct predictions that the application has made and you will be seeing that there are a total of four classes fern madrone toyon and manzanita 20 were incorrectly predicted 80 were correctly predicted so the green ones are predicted correctly while the red ones here are incorrectly predicted all right so the feature of this lobe software is that it is easy to use so the great thing is that there's no code required so if you don't know any computer programming then this is an awesome chance that you could have a look or try at machine learning deep learning and so there's nothing to lose just download it and play around with it and another plus or a good feature of this software is that it is free and it is private meaning that you could download this application and then you could train your own model locally on your own computer and so the data will be safe on your own computer and if you're happy with the application then you could deploy this on the internet using various provider like amazon web service or a store or google cloud gcp so particularly in deploying the model you will be able to export the model as a tensorflow model and then use it locally on your own computer using the python lobe library and so that will allow you to get access to the underlying models that you have already created using the software and so currently the loeb software will allow you to do image classification so you will be able to classify images like for example classifying cats and dogs or the various flowers and in the future they will be releasing new features that you will be able to do with this lobe application particularly including object detection and also tabular data classification so you could give it a try import your own images and then you could label it and then let's see if the application will be able to develop a accurate model to predict it or not and if it can then please share with me in the comments what awesome application are you developing and so in order to build the model you will be needing the image data like it could be an image library already existing on your own computer and then you could import it and then you could label it or if you don't have any images to use for training then you could create a new one using the webcam that you have on your own computer and then you could capture the various images like for example whether you're wearing a mask or not wearing a mask okay so these are some of the example projects that you could try out and let's have a look further in the website so these are some of the platform that you could deploy your models that you have developed here and it includes google apple amazon web service a sewer and also raspberry pi and many more all right so let's have a look at some of the examples here so at the beginning of the video i've already mentioned these tasks that you could do and other would also be like whether the elephant is drinking the water or not right or it could be doing some equipment analytics whether it is safe or not or it could be analyzing the solar system or the galaxy or it could also be analyzing images whether there's forest or no forest okay so these are some of the awesome project ideas that you could create and use your own imagination and create something awesome and let's share with us in the comments of this video and so if you have created something awesome you could share it with all of us in this community and i could also make a video sharing some of your own creation and so if you're finding value in this video please give it a thumbs up subscribe if you haven't yet done so hit on the notification bell in order to be notified of the next video and as always the best way to learn data science is to do data science and please enjoy the journey thank you for watching please like subscribe and share and i'll see you in the next one but in the meantime please check out these videos