CodeProject.AI Version 2.5
- Thread starter MikeLud1
- Start date
nmbgeek
n3wb
If they have an RTX3060 6.2 should offer much better performance right?
Here is another way to try and install.: Run an administrative command prompt and enter
Here is another way to try and install.: Run an administrative command prompt and enter
cd C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2 (assumes default install location for CodeProject) and press enter. Then ../../setup and press enter. That should install the 6.2 via CLI.
Can I uninstall 6.2 without affecting the .NET install? I doesn’t start but maybe i should leave well enough alone.
You can uninstall 6.2 without affecting the .NET install
The model was trained with dark & night images for better night detection, see the below link.
GitHub - OlafenwaMoses/DeepStack_ExDark: A DeepStack custom model for detecting common objects in dark/night images and videos.
A DeepStack custom model for detecting common objects in dark/night images and videos. - OlafenwaMoses/DeepStack_ExDark
github.com
slidermike
Getting the hang of it
Hi guys, been a bit since I posted; life and all.
So I am not finding (I am sure its been discussed) anything current on what (if any) support there is for Intel ARC A750 (the Intel pcie cards) for acceleration.
I need to put BI/CPAI onto an AMD system with an Intel A750 and hoping by now these Intel add-in cards have support similar to the Intel iGPU.
BI/CPAI are on current/latest stable releases 5.9.7.4 & 2.6.5.0
Appreciate the assist.
So I am not finding (I am sure its been discussed) anything current on what (if any) support there is for Intel ARC A750 (the Intel pcie cards) for acceleration.
I need to put BI/CPAI onto an AMD system with an Intel A750 and hoping by now these Intel add-in cards have support similar to the Intel iGPU.
BI/CPAI are on current/latest stable releases 5.9.7.4 & 2.6.5.0
Appreciate the assist.
The model was trained with dark & night images for better night detection, see the below link.
GitHub - OlafenwaMoses/DeepStack_ExDark: A DeepStack custom model for detecting common objects in dark/night images and videos.
A DeepStack custom model for detecting common objects in dark/night images and videos. - OlafenwaMoses/DeepStack_ExDark
Do you use the main stream if available at night, or only for LPR?
My cams are struggling with detecting people at night for sure. Day is fine.
Last edited:
I understand that Chris Maunder and Matthew Dennis are no longer with the Code Project AI Server project.
I am awaiting any word from the other two members of the group as to the future of the project.
Sent from my iPlay_50 using Tapatalk
I am awaiting any word from the other two members of the group as to the future of the project.
Sent from my iPlay_50 using Tapatalk
slidermike
Getting the hang of it
Let us know what you find out. Could be impacting to all the BI users.
Thanks
Chris Maunder and Matthew Dennis are still going to be developing CodeProject.AI
They are moving everything off of CodeProject. Below are some of the new links.
codeproject CodeProject.AI-Server · Discussions
Explore the GitHub Discussions forum for codeproject CodeProject.AI-Server. Discuss code, ask questions & collaborate with the developer community.
github.com
CodeProject.AI Server - CodeProject.AI Server v2.8.1
A Guide to using and developing with CodeProject.AI Server
codeproject.github.io
CodeProject.AI Server - CodeProject.AI Server v2.9.0
A Guide to using and developing with CodeProject.AI Server
codeproject.github.io
Moved to the ipcam-dark model for my cameras at night. Getting multiple person alerts with 80% accuracy now on this plant which I wasn’t before. Anything I can do or would I need to add this as an area to skip?Oh. I was following the Code Project AI Server discussion board at codeproject.com.Chris Maunder and Matthew Dennis are still going to be developing CodeProject.AI
They are moving everything off of CodeProject. Below are some of the new links.
View attachment 204636
View attachment 204637
codeproject CodeProject.AI-Server · Discussions
Explore the GitHub Discussions forum for codeproject CodeProject.AI-Server. Discuss code, ask questions & collaborate with the developer community.CodeProject.AI Server - CodeProject.AI Server v2.8.1
A Guide to using and developing with CodeProject.AI ServerCodeProject.AI Server - CodeProject.AI Server v2.9.0
A Guide to using and developing with CodeProject.AI Server
I must have missed the transition announcement.
No wonder there are no replies there.
My bad.
I will search more carefully next time.
Sent from my iPlay_50 using Tapatalk
slidermike
Getting the hang of it
Similar to PBC I have plants/tree branches that I have excluded due to false alerts but when its heavy rain or heavy winds, they can creep out into the rest of the monitoring window and trigger false again. Its a pain in the butt. To exclude the entire travel path of these limbs would effectively blind me to about 1/3 of the monitoring window. Its a less than stellar solution considering its supposed to have some sort of AI but cant seem to tell a swaying limb from animal/human
Vettester
Getting comfortable
- Feb 5, 2017
- 798
- 814
If masking doesn’t work for you maybe try setting up multiple zones and utilize the “object crosses zones” option in object detections. Another option would be to increase the “min confidence (%)” until you find a happy medium.Similar to PBC I have plants/tree branches that I have excluded due to false alerts but when its heavy rain or heavy winds, they can creep out into the rest of the monitoring window and trigger false again. Its a pain in the butt. To exclude the entire travel path of these limbs would effectively blind me to about 1/3 of the monitoring window. Its a less than stellar solution considering its supposed to have some sort of AI but cant seem to tell a swaying limb from animal/human
slidermike
Getting the hang of it
Thanks @Vettester , the multi-zones and crosses is a good idea. I can give that a try. Hate to raise the confidence as generally it works decent (minus the bad weather events that trigger falses out the wazoo).
Hi,
I am running 2.6.5 and using a single coral. My settings are efficientdet-lite and multi-tpu (though it says its falling back to single tpu).
I often get an error saying unable to run inference at least 1 reference to internal data. Any thoughts on this?
I am running 2.6.5 and using a single coral. My settings are efficientdet-lite and multi-tpu (though it says its falling back to single tpu).
I often get an error saying unable to run inference at least 1 reference to internal data. Any thoughts on this?
That's what happens when multiple threads hit the old single-TPU code at the same time (or it could be a bug in the multi-TPU code, but it sounds like that's not running). I'd be curious as to why it's falling back to single-TPU; fixing that and running the newer multi-TPU code would likely fix the problem.