👁 Computer Vision Cloud Service and Computer Vision 2.0

Remember when we said we have more surprises coming up for Computer vision? ComputerVision V2.0, Cloud, and new OCR. Read all about it 👻. Well, that day has come!

We have two major updates for Computer Vision:

  1. Computer Vision Cloud Service: starting today this is available as part of our on-cloud offering to all customers not only our Community users.
  2. Computer Vision activities version 2.0: a new and improved version of our activity packs that brings a multitude of changes, full details in the documentation and the release notes linked from below.

IMPORTANT NOTE: Starting today, in order to use Computer Vision you will be required to enter an API key, provided to all customers through our Cloud Platform, whether you use Orchestrator on-prem or cloud. Details on how to obtain this can be found in our documentation linked below.

Computer Vision 2.0 Release Notes: https://docs.uipath.com/releasenotes/docs/uipath-ai-computervision-activities#section-v2-0-0
Documentation: https://activities.uipath.com/docs/about-the-ai-computer-vision-activities-pack
Getting an API Key for using the Computer Vision moving forward: https://orchestrator.uipath.com/v2019/docs/asking-for-more-licenses#section-computer-vision-api-key

16 Likes

As I have mentioned in the early post of computer vision activities, It helped me a lot in automating processes in VM. And I hope this version will help me even better

Thanks Team :slight_smile:

3 Likes

Nice feature for the Citrix Automations. Kudos UiPath

2 Likes

@Cosin
Great update team.

3 Likes

Another awesome features again thanks for the update :smiley:

cheers :smiley:

Happy learning :smiley:

2 Likes

Great work guys!!! :smiley::smiley:

2 Likes

great work team UiPath, i look forward to use them !

1 Like

Great work team UiPath… :partying_face::clap:

1 Like

Nice features, Good work Team :clap::clap:

1 Like

Is the computer vision still available to developers? (without enterprise subscription)

I am excited by the possibilities it offers. But it is far from identifying anchors and data points like a human does. I was hoping that CV will make it possible to work on scanned images of AP Invoices also. It looks like it is designed to work with fixed formats appearing in Citrix. It would be great if we can have a combination of find OCR Text and CV. Thereby CV can act like FlexiCapture. Even better to provide for a way to store keywords which can be searched for common data points such as Invoice No, Invoice Date with variants for each column name. Also I found that only Tesserat works well with scanned images, but the text returned is gibberish for small fonts, and also it is very slow. It took couple of minutes to capture two data points from the scanned image.

1 Like

Yes, just create a normal account on cloud.uipath.com and get the key

2 Likes

Yeah… it’s not meant for that, it’s meant for UI automation (no fixed formats tho, it works with any UI). A similar approach is being built for documents but it’s not ready yet.

1 Like

Hello,

Since everyone is allowed to get a key for Computer Vision, how do you differentiate paid customer than community customer?
The Cloud account I created is a community account and seems to have a limit to 30 Megapixels/Min. For a company with a lot of robot, if the same developer are developing all the bot it means same key will be used on all robot and soon enough the threshold will be reached.
Is there any enterprise key that exist ? what are the limit for those keys?

Thank you,
Loïc

Yes, enterprise accounts have different limits: 240Mp/min.

2 Likes