CodeProject.AI Version 2.0

truglo said:
I have doubts it's the cause of delivery not working for you, but you probably should remove all of the "Mark as vehicle" entries since you aren't confirming plates or vehicles on this one.

I vaguely remember seeing those blue/orange double rectangle icons in the ai logs when ai was misbehaving with detecting static objects (especially without alpr:0). I noticed you have that disabled. Maybe enable it to see if it works?
Click to expand...
I will definitely try it. It is funny, you helped my the last time to got it to a point where I was able to recognize most of the delivery trucks. I even documented it, not sure what change since the last time we had it working that I can not get it to work anymore. Thinking aloud, it sees the logo and mark it as 90% accuracy, what else does it wants I will report back after I make those changes either way.

Sent from my SM-S906U using Tapatalk
 
Well, you guys did it, I'm declaring victory yet but yesterday I got amazon truck market as such. Thank you so much for all your help. Next adventure is face recognition, alghouth I do not see too much of it here nowadays. Is because it works or because lot of issues with it.

Thanks again for the help.

Sent from my SM-S906U using Tapatalk
 
  • Like
Reactions: CanCuba
AlphaBlueIris said:
Well, you guys did it, I'm declaring victory yet but yesterday I got amazon truck market as such. Thank you so much for all your help. Next adventure is face recognition, alghouth I do not see too much of it here nowadays. Is because it works or because lot of issues with it.

Thanks again for the help.

Sent from my SM-S906U using Tapatalk
Click to expand...

Most of us have found that facial recognition is more gimmicky and novelty than anything else. If you have to put in 5 or 10 or 15 or 35 pictures or more of yourself in the system for it to recognize it is you...then you shouldn't expect much.... My success rate was under 5% so I moved on to other things LOL.

It can work in certain situations like a business that requires everyone to stop in front of the camera and the camera is at head height. Outside of that, the percentage of being accurate is probably not going to be super high. You will get a lot of false "confirmations" doing a search.

Someone here posted once how horrible it was inside his house identifying his neighbors and others as him. Another guy his kids and wife were being tagged as him.

It is fun to play with, but I wouldn't depend on it for mission critical activities.

Unless you spend the big bucks that casinos and airports have LOL.

Heck even in ideal situations like a business with the camera at ideal height and optimal lighting it fails....

What do you think of Dahua's "Uyghur Warnings"?

Good luck. Wish you no ill will. I'm not here to argue with people simply to contribute when and where I can. This is also followed by lending my views on current technology and how it impacts all of us. It's undisputed that splitting of the atom changed the world but once again man decided to...
ipcamtalk.com ipcamtalk.com
 
  • Like
Reactions: David L
wittaj said:
Most of us have found that facial recognition is more gimmicky and novelty than anything else. If you have to put in 5 or 10 or 15 or 35 pictures or more of yourself in the system for it to recognize it is you...then you shouldn't expect much.... My success rate was under 5% so I moved on to other things LOL.

It can work in certain situations like a business that requires everyone to stop in front of the camera and the camera is at head height. Outside of that, the percentage of being accurate is probably not going to be super high. You will get a lot of false "confirmations" doing a search.

Someone here posted once how horrible it was inside his house identifying his neighbors and others as him. Another guy his kids and wife were being tagged as him.

It is fun to play with, but I wouldn't depend on it for mission critical activities.

Unless you spend the big bucks that casinos and airports have LOL.

Heck even in ideal situations like a business with the camera at ideal height and optimal lighting it fails....

What do you think of Dahua's "Uyghur Warnings"?

Good luck. Wish you no ill will. I'm not here to argue with people simply to contribute when and where I can. This is also followed by lending my views on current technology and how it impacts all of us. It's undisputed that splitting of the atom changed the world but once again man decided to...
ipcamtalk.com ipcamtalk.com
Click to expand...

Yea, im kinda running into the same thing with facial recognition. I have multiple images of myself in different angles, and they all return back to me and percent is high when testing through CPAI. But BI only detects me with the alert maybe 1 out of 50 times. My goal was for it to recognize my roommate and I so that I can set those to not send alert to phones, but keeping "person" alerting to phones. I know under deepstack i had that setup and it had about a 75% success rate, but under CPAI its basically well under 5% success rate. It's not mission critical for me, but would like the ability to do that with CPAI. I just dont know where the breakdown is. Maybe in future versions it would be better, but for now I can only hope this feature gets some attention at some point. Another option I have started to look into is creating a custom model just for my roommate and I and training all the images. It works for other things like animals and such, so why not people as well. I am still reading the documentation on creating custom models..maybe thats a work around? I dunno. but its something I will try at some point. Like I said its not mission critical for me, just a convenience.
 
Last edited:
  • Like
Reactions: David L
Since moving from DS to CP.AI, my BI GUI is very sluggish. Windows open very slowly and camera(s) randomly disconnect/reconnect throughout the day. If I turn on "Use GPU" and reboot, the GUI is even slower and sometimes non-responsive. But, after disabling CP.AI (unchecking "Use AI Server on IP/Port") and rebooting, the problems disappear and BI runs great.

I'm running BI (5.7.05) and CP.AI (2.0.8) on a dedicated PC (i7-4770, 24GB) and never had any performance issues in the past. When CP.AI is on, the BI CPU indicator on the bottom right typically bounces around between 12% and 25%. It occassionally spikes to >90% and then drops back down. RAM is usually 750M-980M. AI average speed is 240ms (6.91/min).

I'm just an end-user so not sure what else to look for or try. Any suggestions?


1679324438927.png1679324510592.png1679324586559.png1679324806594.png1679325113169.png
 
  • Like
Reactions: David L
@Post If you "Use GPU" in BI5, make sure it's enabled on CodeProject.AI side as well.

What Nvidia GPU do you have? If GPU drivers are installed properly and GPU is enabled, it should take care of AI workload so CPU is not bogged down. You should be able to achieve <100ms detection time with decent GPU (when using Mike's models).

Also, for the camera AI screenshot #2, turn off "static object". You shouldn't need it for detecting person.

When "static object" is enabled, BI5 will call CodeProject.AI very often behind the scenes. You can eliminate that workload by turning that option off.
 
Last edited:
Yes, GPU was enabled on the CP.AI side as well but my cards are old! I started with the integrated Intel 4600 and then installed an old GTX 645 that I had laying around. I just tried "Use GPU" again and it still makes the GUI much slower and no difference in the AI detection time. Also, I did not install CUDA -- seemed complicated and not sure it would make a difference given the age of the card.

I also unchecked "Detect/ignore Static Objects" on all cameras and rebooted but unfortunately that didn't help.

1679333962316.png
 
Last edited:
MikeLud1 said:
ONNX version attached.
Click to expand...
Hi Mike. I finally switched from DeepStack to CPAI yesterday. I also added your custom model. Is there anything else i need to do/install? I had the same files as Post, so used your ONIX version.

View attachment 156935

View attachment 156940
Post said:
Since moving from DS to CP.AI, my BI GUI is very sluggish. Windows open very slowly and camera(s) randomly disconnect/reconnect throughout the day. If I turn on "Use GPU" and reboot, the GUI is even slower and sometimes non-responsive. But, after disabling CP.AI (unchecking "Use AI Server on IP/Port") and rebooting, the problems disappear and BI runs great.

I'm running BI (5.7.05) and CP.AI (2.0.8) on a dedicated PC (i7-4770, 24GB) and never had any performance issues in the past. When CP.AI is on, the BI CPU indicator on the bottom right typically bounces around between 12% and 25%. It occassionally spikes to >90% and then drops back down. RAM is usually 750M-980M. AI average speed is 240ms (6.91/min).

I'm just an end-user so not sure what else to look for or try. Any suggestions?


View attachment 157513View attachment 157514View attachment 157515View attachment 157516View attachment 157517
Click to expand...
I have almost have your same PC. Mine is running good on my GPU with four CAMs at 15FPS each.

1679349550625.png
1679349200779.png
1679349244938.png
1679349140972.png
1679349069733.png
1679349414506.png

One thing I did do, but did not need to, just read it helps the CPU and it did drop mine a bit:
Check Limit decoding unless required. What you will notice is when you go to a CAM the video will fast forward to catchup,
took a bit to get use too but it does not affect the recordings...only when you first View.
1679349663203.png

HTH
 
Yeah check your drivers. I read this when i switched over and I am running driver 516.94

1679350206226.png

1679350099720.png
 
Post said:
Yes, GPU was enabled on the CP.AI side as well but my cards are old! I started with the integrated Intel 4600 and then installed an old GTX 645 that I had laying around. I just tried "Use GPU" again and it still makes the GUI much slower and no difference in the AI detection time. Also, I did not install CUDA -- seemed complicated and not sure it would make a difference given the age of the card.

I also unchecked "Detect/ignore Static Objects" on all cameras and rebooted but unfortunately that didn't help.

View attachment 157524
Click to expand...
Did notice your GPU was unchecked.

1679350322573.png1679350462282.png
 
actran said:
@David L Turn off "Default object detection" if you are using Mike's models. And disable "YOLOv5 6.2" and enable "YOLOv5 .NET" if you want to use ONNX models.
Click to expand...
Thanks, I have not tried the custom models yet. But i will make note when I do.

I just recently switched over from DS. So far everything is working good. Probably need to turn off Facial Recon. since I don't use it. It may help processing speed...
 
Post said:
Yes, GPU was enabled on the CP.AI side as well but my cards are old! I started with the integrated Intel 4600 and then installed an old GTX 645 that I had laying around. I just tried "Use GPU" again and it still makes the GUI much slower and no difference in the AI detection time. Also, I did not install CUDA -- seemed complicated and not sure it would make a difference given the age of the card.

I also unchecked "Detect/ignore Static Objects" on all cameras and rebooted but unfortunately that didn't help.

View attachment 157524
Click to expand...
Assuming my 4Gig GPU is what is helping me even though my card is also quite old.

1679351158817.png
 
Post said:
Yes, in that screenshot I was running without the GPU.

Mine is only 1GB. But, even without a GPU, I didn't think CP.AI would bog down BI like it's doing with mine. I ran DS with no problem.
Click to expand...
How many CAMs do you have? Many try using AI only on one at first.
 
  • Like
Reactions: Post
Post said:
Yes, in that screenshot I was running without the GPU.

Mine is only 1GB. But, even without a GPU, I didn't think CP.AI would bog down BI like it's doing with mine. I ran DS with no problem.
Click to expand...

CodeProject has a lot more bells and whistles than Deepstack has and the program is much larger as well, thus needing more horsepower.

Couple that with a 4th gen CPU and GPU with only 1GB and you could run into issues.

I have an older laptop that I use as my test before I do major BI updates. It was the computer I was using all thru when Deepstack came out and I was testing BI versions the moment it came out. It ran Deepstack just fine (albeit times were slow in the low 1,000s) and it won't run CodeProject at all.

You need to start cutting out models and number of cameras and try it one camera at a time until you see it struggle.

Or go back to Deepstack. I am still on Deepstack as there seems to be way too many issues for many folks here for me to risk it on my BI computer.
 
  • Like
Reactions: Post and Nunofya
As an eBay Associate IPCamTalk earns from qualifying purchases.
  • Like
Reactions: Post
Well, things went from bad to worse :facepalm:. Last night I turned off AI on each camera except for one, rebooted, and the GUI was even slower than before. I tried a few reboots, uninstalled/reinstalled CP.AI , and reinstalled the BI update. I let it sit overnight and this morning the status log was full of connection error messages and the GUI was creeping.

Now, I just want to get back to where I was before I started with CP.AI. So, I turned off AI, uninstalled CP.AI, removed the graphics card, re-updated BI to 5.7.05, ran repair database, rebooted the PC, and power cycled the camera POE switch. After all that, the entire PC is very sluggish, including the BI GUI and my cameras are disconnecting/reconnecting every so often. Any suggestions what to try next to get back to normal?
 
  • Like
Reactions: David L
Post said:
Well, things went from bad to worse :facepalm:. Last night I turned off AI on each camera except for one, rebooted, and the GUI was even slower than before. I tried a few reboots, uninstalled/reinstalled CP.AI , and reinstalled the BI update. I let it sit overnight and this morning the status log was full of connection error messages and the GUI was creeping.

Now, I just want to get back to where I was before I started with CP.AI. So, I turned off AI, uninstalled CP.AI, removed the graphics card, re-updated BI to 5.7.05, ran repair database, rebooted the PC, and power cycled the camera POE switch. After all that, the entire PC is very sluggish, including the BI GUI and my cameras are disconnecting/reconnecting every so often. Any suggestions what to try next to get back to normal?
Click to expand...
Make sure you are not running an Anti-Virus...check C: Drive for any memory dump files. Remove any TEMP files. at START Run Enter %TEMP% to get to TEMP directory. Maybe run Maintenance from START. Just to name a few...
 
  • Like
Reactions: Post
You must log in or register to reply here.
Top Bottom