CodeProject.AI Version 2.0

@MikeLud1 Apologies for more questions. Thank you for the great advice above. I have set Nvidia to "maximum performance" in Nvidia Control Panel.

I've stopped YOLOv5 6.2 and started YOLOv5 NET, but can't seem to enable GPU.
I've restarted PC just to be sure and tried again. Nothing in CP.AI server log indicates why GPU enable did not work.YOLOv5 NET.pngserver logs.png
 
@MikeLud1 Apologies for more questions. Thank you for the great advice above. I have set Nvidia to "maximum performance" in Nvidia Control Panel.

I've stopped YOLOv5 6.2 and started YOLOv5 NET, but can't seem to enable GPU.
I've restarted PC just to be sure and tried again. Nothing in CP.AI server log indicates why GPU enable did not work.View attachment 155881View attachment 155882
When you first enable Object Detection (YOLOv5 .NET) it will show CPU after the first detection it should show GPU (DirectML). How is the speed compared to Object Detection (YOLOv5 6.2)
 
Quick question, my car wires are being chew by mice or maybe rabbits, what kind of custom model can I use for it and will codeproject be able to capture those creatures?

Sent from my SM-S906U using Tapatalk
 
Quick question, my car wires are being chew by mice or maybe rabbits, what kind of custom model can I use for it and will codeproject be able to capture those creatures?

Sent from my SM-S906U using Tapatalk
Our Daughter went through this a few years back, they got their two cars in the garage. I know they used some kind of spray. They were hit twice, meaning they got the wires replaced then it happened again. I think their final solution were traps...

 
Our Daughter went through this a few years back, they got their two cars in the garage. I know they used some kind of spray. They were hit twice, meaning they got the wires replaced then it happened again. I think their final solution were traps...

I hear you, unfortunately I have my car outside, no space in the garage. So far I smell like peppermint and my hear are blasted by ultrasonic waves

Sent from my SM-S906U using Tapatalk
 
  • Like
Reactions: David L
I highlighted it in the picture, should I indicate the model in that box to capture the mouse and rabbit? What is the difference between real time images of 4 instead of 10 as an example and is better 500ms or 1 sec?

Thanks for all your help
 
I highlighted it in the picture, should I indicate the model in that box to capture the mouse and rabbit? What is the difference between real time images of 4 instead of 10 as an example and is better 500ms or 1 sec?

Thanks for all your help
 

Attachments

  • 2023-03-06 16_50_15-Blue Iris Remote Desktop - 192.168.0.57 - Remote Desktop Connection.png
    2023-03-06 16_50_15-Blue Iris Remote Desktop - 192.168.0.57 - Remote Desktop Connection.png
    45.2 KB · Views: 80
So now you are getting less then 100ms

@MikeLud1 is there a reason why .NET is so much more performant than 6.2?

Reading all of the guides mention that you should be using 6.2 for GPU and .NET for CPU but it seems .NET works an order of magnitude faster for GPU?
 
I do not see any response, 1) could be a stupid question, feel free to say so or people are on the same boat as I'm. I'm trying to read as much as I can but every screenshot that I see in this forum are different and each has they own settings.

Sent from my SM-S906U using Tapatalk
 
I do not see any response, 1) could be a stupid question, feel free to say so or people are on the same boat as I'm. I'm trying to read as much as I can but every screenshot that I see in this forum are different and each has they own settings.

Sent from my SM-S906U using Tapatalk
No stupid questions.
What works for me, my targets may vary from yours.
Do not use Default Object Detection
Do not use Main stream if available
Use custom models.
I use 100 msec and 4 images, but that is mostly for moving vehicles. Trial and error called for here due to other variables. Like pre-trigger record.

If I were you, I would first experiment using the Codeproject AI explorer. Try the different models, using their samples as well as graphics that you provide. Then take that info to BI.


"An intellectual is someone who can listen to the William Tell Overture and not think of the Lone Ranger."
 
No stupid questions.
What works for me, my targets may vary from yours.
Do not use Default Object Detection
Do not use Main stream if available
Use custom models.
I use 100 msec and 4 images, but that is mostly for moving vehicles. Trial and error called for here due to other variables. Like pre-trigger record.

If I were you, I would first experiment using the Codeproject AI explorer. Try the different models, using their samples as well as graphics that you provide. Then take that info to BI.


"An intellectual is someone who can listen to the William Tell Overture and not think of the Lone Ranger."
That is very nice of you, great tips. Thanks. Things are moving so fast in the blueiris codeproject world that is hard to follow.

Sent from my SM-S906U using Tapatalk