-
Hello, does the given code have to be trained and tested on GPU
-
HI, thanks for your great work. I want to use the pretrained model on PKL dataset to test a video, how should I do step by step?
How to save the model and how to input a video for inference?
-
-
Hello Teoge,
I want to add on more type of OPEN Slot parking type, following steps done
1) Labelled new type of Open SLOT Type with index number "2"....
Now "0"---> for T Slot, "1"--->for L Sha…
-
Hello,
Is it possible to use detectParking on a picture of a parking lot instead of the video?
If so, could you please give me a guideline how to do so?
Thanks,
-
- Input: EnvironmentModel.msg, ParkingInProgress.srv
- Output: TrajectoryMetaInput.msg, ParkingInProgress.srv
Features:
- [x] Process environment data, measure parking spot sizes
- [x] Detect/Flag…
-
Hi , I m intersted in the implementation and could you able to share your notebook for seeing how its performs on other images too? I installed all dependencies but still no luck Unfortunately I m…
-
Hi,
kann mir jemand sagen, ob man in dem WiFi-M.2-Port auch eine Google Coral TPU nutzen könnte?
z.B.
[CORAL M.2 ACCELERATOR MIT DUAL EDGE TPU](https://www.pollin.de/p/coral-m-2-accelerator-mit-d…
-
### Search before asking
- [X] I have searched the YOLOv8 [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussions) and fou…
-
First, I want to thank you for using your skills in a way to help other in these trying times. Developing new technology to help others is a laudable endeavor. That said, I have some concerns about th…