General questions

I use the offline sdk, but the recognition is very slow after integration. Using your demo, it’s very fast

Hello, I have trouble integrating your offline sdk.

 All points of my notes are saved in a file. When recognition is needed, I will read out all the points, then call editor.pointerDown, pointerMove, pointerUp and other methods to pass the data, and finally call the convert() method. But the recognition time is very long. When I use your demo, there is basically no need to wait for recognition. Please tell me how to solve this situation


Thank you for your update.

Please share with us more details on your integration:

how do you capture the strokes? Do you render them with the SDK?

Which content part are you using? Which editor?

Best regards,


Thank you for your reply. 

I only need to get the recognition results and render the recognition results using the sdk. 

In addition, I would like to ask if there is any way to arbitrarily edit the label of the recognition result jiik data. I found that I can only edit one of the candidate words. If all the candidate words are incorrect, then no matter how I modify it, it will be wrong.

Dear gw_2601,

Thank you for contacting us.

The recognition speed depends on the quantity of strokes that you are sending.

Also depending on whether you use our SDK for rendering the strokes and whether you want to use interactivity  ,integration should be different.

Sharing with us more information about your use case would help us better advise you.

Best regards


Login or Signup to post a comment