Our team has requested to check whether we can create generic automation using AI Computer Vision. Can we develop a single automation solution for different web applications?
Suppose two screens in the attached image are from two different web applications. I need to create a single automation solution to fetch details such as Name, Age, and Address. There may be more applications.
Is this achievable using AI Computer Vision? If it is achievable, can anyone please help me with an example solution?
it can work with little changes but might not go with one single..as it uses a backend ml model too which would try to identify based on position and text and relative visual representation as well
I have few more queries related to AI Computer Vision. Thanks in advance!
Can we make a dynamic solution for the application in below screenshot using AI Computer Vision? Is there any option to pass dynamic values for Currency and return Amount and Change using AI Computer Vision? This might be possible using usual web scraping methods, but this query is specific to AI Computer Vision.
Ideally web scrapping would do but not sure why youwant to use ai only
One way you can do is using kind og anchor and add variable in text in descriptor and also need to paas coordinates too by finding the relative co ordinates of the currency you found
Thank you all for the valuable information and help!
I am new to UiPath, so my questions might be irrelevant.
Can anyone please help me understand more about Computer Vision by answering the following questions?
If an application can be fully automated using both normal web scraping methods and Computer Vision, is there any benefit in using Computer Vision over normal web scraping?
Can you provide a scenario where Computer Vision can be used over normal web scraping to automate an application, even if it is not in a Citrix or remote environment?
If the application frequently has slight changes in the colour or position of elements, which method would be more beneficial: normal web scraping or Computer Vision?
Additionally, it would be great if you could add more points to help me understand the benefits of using Computer Vision in a native system (not in Citrix or remote environments).
Computer vision is more tolerant than web scraping methods. If selectors change computer vision is more stable. Therefore, I would even often recommend to use computer vision.
Further, it is easier to automate with computer vision than identifying fitting selectors. Finding the right selector can be tricky sometimes.
The field also shifts more and more to Computer vision. UiPath has introduced a self-healing agent, which is aligned with computer vision.
If the position changes than computer vision is better. If the colour changes both will work, if you do not name the colour in the web scrapping selector.
