[tool] [tutorial] Free AI Person Detection for Blue Iris

barnyard

n3wb
Joined
Aug 9, 2020
Messages
22
Reaction score
5
Location
United States

Use the lower links and note they are going over to docker only I believe ;

Run
deepquestai/deepstack:latest
or
deepquestai/deepstack:cpu-x5-beta

It's much much faster. :) :) :)

View attachment 73095
So, how's the accuracy on the new version? I'm constantly getting person notifications on this cat that's been hanging around the house. Also, when they say that noavx is supported out of the box, do I still add the noavx tag to the command when I run it?

Sent from my Pixel 3 using Tapatalk
 

Village Guy

Getting the hang of it
Joined
May 6, 2020
Messages
194
Reaction score
85
Location
UK
:confused: that download is in my post already.
My post was intended to advise that their 'latest' version may need updating, hence I gave a time reference. Sorry if you felt it was simply duplication of your earlier helpful post.
 

Tinbum

Pulling my weight
Joined
Sep 5, 2017
Messages
420
Reaction score
111
Location
UK
So, how's the accuracy on the new version? I'm constantly getting person notifications on this cat that's been hanging around the house. Also, when they say that noavx is supported out of the box, do I still add the noavx tag to the command when I run it?

Sent from my Pixel 3 using Tapatalk
It now give a proper % not just between 89 and 100.
 

MnM

Young grasshopper
Joined
May 14, 2014
Messages
86
Reaction score
16

Use the lower links and note they are going over to docker only I believe ;

Run
deepquestai/deepstack:latest
or
deepquestai/deepstack:cpu-x5-beta

It's much much faster. :) :) :)

View attachment 73095
Which one is much faster? deepstack:latest or deepstack:cpu-x5-beta?
 

tripp396

Young grasshopper
Joined
Jun 18, 2020
Messages
42
Reaction score
11
Location
Minnesota
Got a quick question on building this from source. I was able to build it in Visual Studio fine and it works great. Is there an easy way to save my options and camera setup from build to build?
 

Senor Pibb

Young grasshopper
Joined
May 22, 2020
Messages
43
Reaction score
18
Location
Greer, SC
In the Status tab, is there a way to display the labels for the colors? I can't remember what each color is from the org app.
 

tripp396

Young grasshopper
Joined
Jun 18, 2020
Messages
42
Reaction score
11
Location
Minnesota
Another question: Is there a way to just send through the object detected without the confidence level? For example I am using the [Detection] variable in an MQTT topic and while I though that used to just send through something like "Person", it seems to send through "Person (95%)". Wondering if there is a way for just the object to have a variable.
 

Tinbum

Pulling my weight
Joined
Sep 5, 2017
Messages
420
Reaction score
111
Location
UK
In the Status tab, is there a way to display the labels for the colors? I can't remember what each color is from the org app.
I now copy the files from the debug folder and paste in the AITool folder without deleting files first. keep backups.

You need to make sure the Camera Folder, JSON and SQLITE files don't get deleted.
 

Senor Pibb

Young grasshopper
Joined
May 22, 2020
Messages
43
Reaction score
18
Location
Greer, SC
I now copy the files from the debug folder and paste in the AITool folder without deleting files first. keep backups.

You need to make sure the Camera Folder, JSON and SQLITE files don't get deleted.
I am sure I am missing something. The labels are stored in the JSON or SQLITE? I can go back and look for them. I do the following: make backup of current folder, copy new files from zip of new ai tool to current folder. attempt to run and test. restore if needed. But I must have missed something. can you suggest where the labels are stored for the status tab.?
 

MnM

Young grasshopper
Joined
May 14, 2014
Messages
86
Reaction score
16
I am using a separate virtual Ubuntu for deepstack

I login to my Ubuntu server and issue this command (replace deepsatckai with your container name)

sudo docker logs -f deepstackai

I posted a few pages back my configuration:

 

tripp396

Young grasshopper
Joined
Jun 18, 2020
Messages
42
Reaction score
11
Location
Minnesota
I am using a separate virtual Ubuntu for deepstack

I login to my Ubuntu server and issue this command (replace deepsatckai with your container name)

sudo docker logs -f deepstackai

I posted a few pages back my configuration:

That is interesting. I'm using a similar setup and my times are close to a full second. Half that when using deepstack right there on the windows server
 

MnM

Young grasshopper
Joined
May 14, 2014
Messages
86
Reaction score
16
I believe is all to do with the individual components.
I use Samsung 970 Pro NVMe M.2 1Tb for my virtual servers (OS drives) and spinning disks (7200prm) for Data drives. It gives my next to 0 latency.
 

tripp396

Young grasshopper
Joined
Jun 18, 2020
Messages
42
Reaction score
11
Location
Minnesota
I believe is all to do with the individual components.
I use Samsung 970 Pro NVMe M.2 1Tb for my servers. It gives my next to 0 latency.
Makes sense, just running off spinning drives right now for my server.
 

MnM

Young grasshopper
Joined
May 14, 2014
Messages
86
Reaction score
16
My BlueIris server is like - using the Samsung NVME for OS, Intel DC S3710 800GB SSD (write intensive it will last me a minimum of 15 years) for 1st drive where BI records 24/7 and generates the alert files (it also has the BI DB and files) - then I have spinning drive where I archive BI recordings and alerts to. Again next to 0 latency for BI virtual server.
 

Village Guy

Getting the hang of it
Joined
May 6, 2020
Messages
194
Reaction score
85
Location
UK
If you don't mind helping a newbie ... how / where are you getting this data from? I am running the docker version on Unraid.
Just install the latest version on deepstack and you may experience near realtime analysis . Check out my earlier post.
 

Tinbum

Pulling my weight
Joined
Sep 5, 2017
Messages
420
Reaction score
111
Location
UK
I am sure I am missing something. The labels are stored in the JSON or SQLITE? I can go back and look for them. I do the following: make backup of current folder, copy new files from zip of new ai tool to current folder. attempt to run and test. restore if needed. But I must have missed something. can you suggest where the labels are stored for the status tab.?
I was replying to visual studio method not the zip file but all the settings are in;

AITool.Settings.JSON
 
Top