5.5.8 - June 13, 2022 - Code Project’s SenseAI Version 1 - See V2 here https://ipcamtalk.com/threads/codeproject-ai-version-2-0.68030/

Thanks for the response. I have 13 cameras going and plan on adding more (cameras still in boxes). I thinking you may be right about the GPU. But wondering if it could be slow internet on my part. And it’s slooooooooooow. I did notice that the CPU really jumps up when I run AI analytics.

BI and SenseAI are not going thru the internet to perform the functions - that is made up of your CPU and GPU performance.

Having too many cameras using SenseAI and not a powerful enough CPU and GPU can impact performance and speed.

The bigger question is do you need SenseAI for every camera? If you have cameras with built-in AI, in many use cases that is sufficient to use and not use SenseAI.

I have found the Dahua IVS to be better, as have many of us here, especially if all you need is human or vehicle detection.

Depending on your computer, you may experience a significant lag, delay, or missing of the AI using SenseAI.

I have tested all different ways and in my experience, using the camera AI and ONVIF trigger works well for my case. Now of course using the camera AI you do not get the nice rectangle around the object.

I do use BI AI for a few cameras, but maybe you should look at not using it for every camera if the lag/delay is too much for you.

As with most things, YMMV.

 
I'm still learning and can't help you but I believe in getting things today that will still work tomorrow so I bought a much bigger one then I actually need. I have a GTX 1650 mini super from e-bay for just over $100. I have no idea what I'm going to need as AI gets bigger and faster and if all else fails I can get a new computer and move the GPU to the new one because the GPU is not that old.
There are a lot of very smart people here that can help you by far better then I.
Fred
 
  • Like
Reactions: 1fxman
BI and SenseAI are not going thru the internet to perform the functions - that is made up of your CPU and GPU performance.

Having too many cameras using SenseAI and not a powerful enough CPU and GPU can impact performance and speed.

The bigger question is do you need SenseAI for every camera? If you have cameras with built-in AI, in many use cases that is sufficient to use and not use SenseAI.

I have found the Dahua IVS to be better, as have many of us here, especially if all you need is human or vehicle detection.

Depending on your computer, you may experience a significant lag, delay, or missing of the AI using SenseAI.

I have tested all different ways and in my experience, using the camera AI and ONVIF trigger works well for my case. Now of course using the camera AI you do not get the nice rectangle around the object.

I do use BI AI for a few cameras, but maybe you should look at not using it for every camera if the lag/delay is too much for you.

As with most things, YMMV.

I appreciate your reply. I was wondering about what you said about just using the cameras AI. And Just yesterday I started doing just that with 2 of my AI cameras. Thanks for that suggestion. My computer is i7-8777 or something like that. I’m not at the computer at the moment so I’m to sure about Ram. But I think it has 250 SSD and I added a 2TB purple hard drive. Just wondering what you meant about “BI AI”? Is that the same as CP AI? Thanks.
 
I appreciate your reply. I was wondering about what you said about just using the cameras AI. And Just yesterday I started doing just that with 2 of my AI cameras. Thanks for that suggestion. My computer is i7-8777 or something like that. I’m not at the computer at the moment so I’m to sure about Ram. But I think it has 250 SSD and I added a 2TB purple hard drive. Just wondering what you meant about “BI AI”? Is that the same as CP AI? Thanks.

An i7 8th gen should be sufficient, but it all depends on how many cameras and how many triggering at the same time as to the speed and capability.

I used the term "BI AI" as a general term meaning using DeepSack or CodeProject or any of the 3rd Party AI tools folks are using with BI.

It is certainly fun to use the AI with BI and get all those orange rectangles and stuff, but after playing with it, for my needs, I want speed and efficiency and to be notified of the important stuff. I only care about people or cars, so all the other items in the BI AI aren't really of any need for me.

I have tested my camera field of views with IVS versus Deepstack and in most instances, using the camera AI has resulted in better success. Of course YMMV and it really is dependent on the field of view.

It is worth repeating that every situation and Field of View is different. What works for some doesn't for others.

But I still use BI AI on a few cams.

For example, for a tightly zoomed in camera like a Z12E for plate reading duty (LPR), the camera AI sucks as the camera needs time to identify the object, determine if it is something you want, compare it to your IVS rule and then decide to trigger or not. These tight views and it will miss them, especially at night.

I run a few cameras with DS because for that particular field of view, it was providing better results.

Here is an example. At night the camera AI would struggle with this tight view. It has a straight on angle of the street to get a side profile of a car and it would miss it a lot of times because the vehicle just isn't in the field of view long enough, so this is a great candidate for DS.

Now the issue I had with DS is that it would either find a car but the alert image would be the lightshine on the street or just a part of the vehicle, or it would trigger out nothing found due to headlight bounce off the street.

1670007802889.png

BI AI has a "to cancel" option, which means it will analyze EVERY image to determine if the item is in it. Once I added a cancel banana in the field, it now will go thru all the images and select the best one, which gives me the whole vehicle in the frame and it eliminated the nothing found as well. It makes for scrubbing video much quicker as I can skip looking at video of known vehicles.

1670007840044.png

So even though I am a fan of camera AI, there are instances where BI AI will be better, or even BI motion detection. I use each one of them depending on the field of view and what is to be accomplished with that field of view.
 
An i7 8th gen should be sufficient, but it all depends on how many cameras and how many triggering at the same time as to the speed and capability.

I used the term "BI AI" as a general term meaning using DeepSack or CodeProject or any of the 3rd Party AI tools folks are using with BI.

It is certainly fun to use the AI with BI and get all those orange rectangles and stuff, but after playing with it, for my needs, I want speed and efficiency and to be notified of the important stuff. I only care about people or cars, so all the other items in the BI AI aren't really of any need for me.

I have tested my camera field of views with IVS versus Deepstack and in most instances, using the camera AI has resulted in better success. Of course YMMV and it really is dependent on the field of view.

It is worth repeating that every situation and Field of View is different. What works for some doesn't for others.

But I still use BI AI on a few cams.

For example, for a tightly zoomed in camera like a Z12E for plate reading duty (LPR), the camera AI sucks as the camera needs time to identify the object, determine if it is something you want, compare it to your IVS rule and then decide to trigger or not. These tight views and it will miss them, especially at night.

I run a few cameras with DS because for that particular field of view, it was providing better results.

Here is an example. At night the camera AI would struggle with this tight view. It has a straight on angle of the street to get a side profile of a car and it would miss it a lot of times because the vehicle just isn't in the field of view long enough, so this is a great candidate for DS.

Now the issue I had with DS is that it would either find a car but the alert image would be the lightshine on the street or just a part of the vehicle, or it would trigger out nothing found due to headlight bounce off the street.

View attachment 147236

BI AI has a "to cancel" option, which means it will analyze EVERY image to determine if the item is in it. Once I added a cancel banana in the field, it now will go thru all the images and select the best one, which gives me the whole vehicle in the frame and it eliminated the nothing found as well. It makes for scrubbing video much quicker as I can skip looking at video of known vehicles.

View attachment 147238

So even though I am a fan of camera AI, there are instances where BI AI will be better, or even BI motion detection. I use each one of them depending on the field of view and what is to be accomplished with that field of view.
Thanks for the info. a lot to learn. Another question, if I'm using camera AI, how do I get a recorded clip? the camera is picking me up. BI is sending push notification but no clip recorded. what say you?
 
Nope you set up BI to a accept ONVIF triggers from the camera and BI does the alerting no different than you do now...

Go into the camera and set up smart plan with IVS, then go to the IVS screen and draw IVS rules (tripwire or intrusion box) and then select the AI you want it to trigger on (human or vehicle).

Then in BI, there are a few places you need to set this up in BI (assuming you already set up the IVS rules in the camera GUI):

  • Go in to BI and select Add Camera
  • Type in the IP address, username, and password
  • Hit the Find/inspect button and let BI choose the right settings.
  • In Camera configure setting check the box "Get ONVIF triggers".
  • Hit Find/Inspect on the camera setting to pull the coding for the triggers.
  • Go into Motion Setting and select the "Cameras digital input" box.
  • On the Alerts tab uncheck the Motions Zones tab (those are alerting you to any BI motion in those areas)
  • On the alerts tab set up how to be notified
 
Nope you set up BI to a accept ONVIF triggers from the camera and BI does the alerting no different than you do now...

Go into the camera and set up smart plan with IVS, then go to the IVS screen and draw IVS rules (tripwire or intrusion box) and then select the AI you want it to trigger on (human or vehicle).

Then in BI, there are a few places you need to set this up in BI (assuming you already set up the IVS rules in the camera GUI):

  • Go in to BI and select Add Camera
  • Type in the IP address, username, and password
  • Hit the Find/inspect button and let BI choose the right settings.
  • In Camera configure setting check the box "Get ONVIF triggers".
  • Hit Find/Inspect on the camera setting to pull the coding for the triggers.
  • Go into Motion Setting and select the "Cameras digital input" box.
  • On the Alerts tab uncheck the Motions Zones tab (those are alerting you to any BI motion in those areas)
  • On the alerts tab set up how to be notified
Thanks. I have all my cameras setup. But I usually use Dahua settings so as to use the audio capability. Do I have to use "find/inspect"? I also do have "Get onvif triggers" checked. I also have all AI IVS Smart Plan setup.
 
If your cameras are two way audio then yeah you need to use however you set it up. If they are like the 5442 series that is just a mic, then the find/inspect and it finding generic is fine.
 
  • Like
Reactions: 1fxman
Have you tried using the yolov5l model? It’s not 100% accurate, but I have found that it works well detecting our dog.
Hi - thanks for this. I finally got around to trying it today, and while it is still not 100% reliable, it is nevertheless far more accurate than the other models I tried (at least, for cats). It is pretty good at detecting people though.
 
  • Like
Reactions: Vettester
I tried Mike Lud's package detection model. Works great! Initially I was using an Outlook email for the alerts, which was working for awhile. Then suddenly the email alerts stopped coming. I went back into the email server settings and ran a Test, and got the below error.

I switched to Gmail and it works every single time. Any idea how to resolve the Outlook error? Or is it just, "don't use Outlook."

test-blue-iris-edit-email-server.png
test-configure-email-alert.png
email-error.png
 
I tried Mike Lud's package detection model. Works great! Initially I was using an Outlook email for the alerts, which was working for awhile. Then suddenly the email alerts stopped coming. I went back into the email server settings and ran a Test, and got the below error.

I switched to Gmail and it works every single time. Any idea how to resolve the Outlook error? Or is it just, "don't use Outlook."

View attachment 147467
View attachment 147468
View attachment 147466
Looks like you exceeded the amount of emails you can send - max amount of free emails in a defined period. "RefuseQuota" and "ShowTierUpgrade" aka pay to increase your limits is what that means to me. This would be extremely common so as to avoid turning that mail server into a spam mailer

 
Looks like you exceeded the amount of emails you can send - max amount of free emails in a defined period. "RefuseQuota" and "ShowTierUpgrade" aka pay to increase your limits is what that means to me. This would be extremely common so as to avoid turning that mail server into a spam mailer

Fascinating. Thank you kindly!
 
I just got one of the new Annke ncd800 180 degree camera, Empiretech has a new one also

Any support or effort towards support in code project? regarding the super wide pictures it has

would ai just scan a super super narrow wrong aspect ratio pic?
 
I just got one of the new Annke ncd800 180 degree camera, Empiretech has a new one also

Any support or effort towards support in code project? regarding the super wide pictures it has

would ai just scan a super super narrow wrong aspect ratio pic?
I have an hik 180 colorvu. I have no issues using code ai with this camera