argumentum Posted September 1 Share Posted September 1 1 hour ago, yutijang said: Hello, thank you for sharing this awesome UDF! He hasn't log in in 4 years now, but I'll thank you in his behalf 😇 1 hour ago, yutijang said: How can I train it to recognize new objects, specifically the ones I designate for the UDF to detect? Well, that part is unrelated to the UDF ( that I've tested back in 2020 but never used ). That has to do with the project he left links for on the 1st post. Do go to those links and find out from those references. That is your best bet. yutijang 1 Follow the link to my code contribution ( and other things too ). FAQ - Please Read Before Posting. Link to comment Share on other sites More sharing options...
yutijang Posted September 1 Share Posted September 1 (edited) @argumentum Thank you for the quick response! After further research, it seems that a machine learning process is required to either add or create new content for the objects in the .weights file. Edited September 1 by yutijang argumentum 1 Link to comment Share on other sites More sharing options...
sloppyjoe Posted September 2 Share Posted September 2 I have done this on my PC, creating a custom weights file. Youtube has some good resources on how to start this process, look for yolo and darknet training. It will take many hours of research, possibly a weekend of learning depending how quick you can pick it up and if your PC can support the training process. I have a RTX 3070 by the way and got it all to work. I was able to install everything needed for free, and using a program called labelimg, also free, start to build a library of labeled images to use to train on. Good Luck with the process. yutijang 1 Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now