Hello guys, welcome to the 14th session of using Microsoft Word framework Luis and cognitive services tutorial for beginners. In this session we will learn what is computer vision and set up the environment for it. In our example, it is a part of cognitive services. It is used to extract information from images to categorize and process visual data. From the image we can obtain text, description and domain specific models to identify content and label it with confidence. Use added arrestee settings to enable automatic restriction of added content.
It would be very useful if you allow users to upload content to your site. We can also identify image types and color schemes in features. We can retrieve tax from image which is known as OCR, optical character recognition I usually use this to obtain text from images that contain costs. Even we can also retrieve text from handwritten text with an image. We can also identify celebrities or some popular places. Sometimes we need to store thumbnail for the image.
We can use cognitive services to generate those thumbnails. Computer Vision API now supporting videos along with image. However, in this series, I will show you how to work with images. It is an example of computer vision. For this image description is people waiting at a train station. It contains tracks like a train, platform, station, subway, etc.
The image format is GPG. Here are its dimensions clipart can be between zero to three And those flags represent whether the images joined back in by contains added arrestee contents and the main colors within the image like a switch to the studio and integrate those with both fingers first of all we need to add Microsoft dot project reason Microsoft dot Azure Cognitive Services definition packages from end up now as your cognitive services packages under development, so he does not contain models as is called returning text from the image. So I will use project expert for only text recognization. They can, you may pause the video, that package might be publicly released and contains classes for tax recognization. So use these instead of project of score Good I have created a new dialog. It contains an operation that exists in a status thing we will wait for a message received.
Within that displayed dialog, after selecting value, set it to a variable that we have created and ask the user to upload attachments. After these we have retrieved my images settings will upload and wait for message received. In those constant variables, we need to set Computer Vision API key and endpoint. Let us create a play for computer vision. Click on New such vision select Computer Vision API. Click on Create I'm gonna need APA select the location that you want to host and remaining Australia is a country I have already created so opening that is click on overview is endpoint URL that is updated in our project.
It was hot in Australia is selected. Now click on show access please copy any of those keys. Here I have created a class for this In which I just added necessary packages that were required for Vision API. Then we'll start with operating a description from the image. Create a variable name, and description, image, description of image, and added in the option list. After receiving an image, we need to perform an operation based on the option that user selected from from dialog.
The user may add multiple images, so we need to look through each image and often description use Cognitive image and past key region and image. We have not stated this matters. So let us create the parameter names and make it a thing. Here we need to create competence issue API mistakes API key service client credential and content rather often description we just need to call describe images think matter. Somebody says we need to retry the first capture and update text. Data, save it and publish it.
And tested in the emulator have downloaded some images from being to face Vision API. And we got an elephant that is standing in the dark. Let us test with other images too. Now I'm going to select this photo we got a plate of food a little slice of pizza. Now I'm going to select this photo got Emma Watson looking at the camera. There's a play smart enough to identify some celebrities and places.
Now I'm going to opt in faces within images. For this copy, this method can change me. Here we need to use analyze images, think methods in which we need to pass a list of things that Want to retry from here we just need faces. Okay image analysis using analyze image a sync method, pass image URLs from the quite features. Here I will store all the details in the variable. faces property of image analysis contains a list of face description.
I'm adding that found total count faces in image and look to each crazy for each face display extended And eat. They can eat as a string. Within dialogue let us create one more constant for faces in English display in Chrome dialog. Let's copy this case. Change the name of case and variable and update my attorney. Let us save all files and publishing Here I'm going to send an image of the record one face or for me with age 52.
Select this image, which contains multiple faces, got three faces, all enemies and their ages 4929 and 30 respectively. Now let us see what we have learned in this session, we saw where we can use Computer Vision API. In the dialog, we have created some constant variables. Then we have used by creating compiler and in switch case for compiling, then factories and API and endpoint and reaching. In hand first, use describe images the method to return description for image. And use analyze image a simple matter to retry image analysis, in which we need to pass a list of initial feature types.
Here we have just returned faces. Then we use this anchor in our dialogue. We just need to pass content URL for each attachment that is set up. In this session, we saw how to retry description and face. In next session, we will continue to reject APA to retry text and some other details of images. Thanks for watching this video.
If you have any doubts, please feel free to contact me. Have a nice day.