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

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
291
Reaction score
161
Location
UK
Ok- I know there is a guide of sorts on page 1 about using Docker. But if you are like me it is very hard to know how to do this because I am not that smart in all this stuff I see - Docker, Docker for windows, VM, Ubuntu, noavx, Linux, container, portainer and on and on. I just want to keep using my windows set up until it stops working then I guess I'm out, because I have no idea how to switch this all over to using "DOCKER" do I have to delete the current DQAI stuff off the computer? Do I just install it all over the top of everything, I run everything of a separate WIN 10 pro box, can I still do this with DOCKER? Do I really have to run this part on 1 computer and the other part on another computer? I know I sound like a whinny little BIAAA#$CH and I am sorry but for a real novice it is very very frustrating. Can anyone guide me to a real Barney style example of how to do this since we know the windows version is outta here. I really would like to keep up since i truly enjoy all the super hard work everyone has put into this. Thanks.
First and foremost don't think you need special skills! I have been using deepstack running under docker almost from the beginning because it started automatically with windows and was easy to update versions. Just think of docker as an intermediary program that translates everything. If you check out the link on the original post that points to the docker install and come back with questions I will do my best to help. First install cocker just like any other program an we can take it from there
 

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
291
Reaction score
161
Location
UK
@balucanb Sorry I should have said install Docker from the windows install program NOT the install referenced at the begining of this thread as it may be confusing. DOCKER will install stand alone. Get back to us when you have it installed and we can then get you up and running deepstack.
 

balucanb

Getting the hang of it
Joined
Sep 19, 2020
Messages
146
Reaction score
23
Location
TX
First and foremost don't think you need special skills! I have been using deepstack running under docker almost from the beginning because it started automatically with windows and was easy to update versions. Just think of docker as an intermediary program that translates everything. If you check out the link on the original post that points to the docker install and come back with questions I will do my best to help. First install cocker just like any other program an we can take it from there
Thanks Man! I really appreciate it. Just been one of those weeks for me, Kid getting married tomorrow, People at the day job are morons, My POS Smart things hub went dark for 9 freaking hours with no apparent reason and then I thought i was screwed with DQAI! I'm just going to do some day drinking.. Scotch at 12 am..that cool right? LOL.
 

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
291
Reaction score
161
Location
UK
Thanks Man! I really appreciate it. Just been one of those weeks for me, Kid getting married tomorrow, People at the day job are morons, My POS Smart things hub went dark for 9 freaking hours with no apparent reason and then I thought i was screwed with DQAI! I'm just going to do some day drinking.. Scotch at 12 am..that cool right? LOL.
Seems to me you have earned that drink. Enjoy
DQA will still be here tomorrow
 

tripp396

Getting the hang of it
Joined
Jun 18, 2020
Messages
65
Reaction score
30
Location
Minnesota
Was able to get the new docker container up and running on the windows host that blue iris and the aitool reside along with throwing everything on the extra SSD i had laying around. Getting around 250ms on the deepstack side vs 550ms using the normal windows server. This is all over 950ms using the docker container on a linux host somewhere else on the network.

Brings average time to around 500-600 ms for the whole thing which is great!
 

Senor Pibb

Getting the hang of it
Joined
May 22, 2020
Messages
77
Reaction score
36
Location
Greer, SC
Just install the latest version on deepstack and you may experience near realtime analysis . Check out my earlier post.
Hi... yes, have the latest running. But I wanted to know where to get the "log" our output of the execution time from post 21101603394252173.png
 

Senor Pibb

Getting the hang of it
Joined
May 22, 2020
Messages
77
Reaction score
36
Location
Greer, SC
May I suggest the following: Not that smart of guy either.

I did the following:
  1. Watched this video several times -
  2. Then watched it again as I setup BI and configured the original GentalPumpkin software and Deepstack.
  3. I run Unraid and within in unRaid you can run dockers. Its a very simple process to enter in the URL to the repository of the docker and configure a few iP address options. But the same can be said for Ubuntu docker. Portainer is only needed if you want to simplify the administration of dockers / adding dockers as you add more and more dockers. If you are only going to run the one deepstack docker then I don't really see the need for portainer.
  4. when I pulled the docker it was already activated. did not have to do anything. (deepquestai/deepstack:latest) is the repository I use. I use mode=high and VISION-DETECTION = True
  5. After you have this working as desired, then move up to the latest work stored at (very nice fork of the GentalPumpkin) Releases · VorlonCD/bi-aidetection
  6. Last, I have no problem attempting to help while using google meetings if needed.
Again, I am not that good at it either, but with the help of others and that video / this forum I have it working well.
 

MrPie

n3wb
Joined
Oct 6, 2020
Messages
11
Reaction score
0
Location
chile
Hello. This have been working very well for me and I got some weird ideas but I need a little help.

I want to make my dog go to my entry entry door if a cat or a human appears there. Many times he is far away and is not able to detect movement there.
I can train him to go where a certain sound is made, so I'm thinking on using this program to make that sound each time a cat is detected.
I'm thinking on using a speaker, or even better something that can play infrasounds so I can play it loudly.
I see that this program can trigger a certain URL, so the question is: Is it possible to configure a speaker to make a sound when a url is visited? That speaker would be placed in more than one place.

I know nothing about home automation so I'm lost but I imagine that there must be a program or server that does exactly that.
Hopefully someone can help.

Thanks
 

pbc

Getting comfortable
Joined
Jul 11, 2014
Messages
1,024
Reaction score
156
Out of the blue (seemingly anyhow) yesterday I started getting these errors which I just noticed today.

[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!
[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!
[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!

Any idea what is causing it? I'm still getting notifications.
 

tripp396

Getting the hang of it
Joined
Jun 18, 2020
Messages
65
Reaction score
30
Location
Minnesota
What variable can I use to get just a single detection? If I just want the variable to say "Person" or "Car" what variable do i use?

Using [Detection] gives me "Person (99%)" which seems to be the same as summary.
 

blargrx

n3wb
Joined
Oct 28, 2018
Messages
13
Reaction score
11
Location
Stockton, CA
Hello! Just found this awesome project. As someone just starting all of this, Is there an updated guide somewhere? Do you just follow GP guide on page 1? I've tried to read over the most recent 25 pages but it's quite a bit more technical than I'm able to wrap my head around.
May I suggest the following: Not that smart of guy either.

I did the following:
  1. Watched this video several times -
  2. Then watched it again as I setup BI and configured the original GentalPumpkin software and Deepstack.
  3. I run Unraid and within in unRaid you can run dockers. Its a very simple process to enter in the URL to the repository of the docker and configure a few iP address options. But the same can be said for Ubuntu docker. Portainer is only needed if you want to simplify the administration of dockers / adding dockers as you add more and more dockers. If you are only going to run the one deepstack docker then I don't really see the need for portainer.
  4. when I pulled the docker it was already activated. did not have to do anything. (deepquestai/deepstack:latest) is the repository I use. I use mode=high and VISION-DETECTION = True
  5. After you have this working as desired, then move up to the latest work stored at (very nice fork of the GentalPumpkin) Releases · VorlonCD/bi-aidetection
  6. Last, I have no problem attempting to help while using google meetings if needed.
Again, I am not that good at it either, but with the help of others and that video / this forum I have it working well.
perfect, that's what I needed to know! I'll start tomorrow working down your steps
 
Joined
Oct 1, 2020
Messages
18
Reaction score
2
Location
Texas
Was able to get the new docker container up and running on the windows host that blue iris and the aitool reside along with throwing everything on the extra SSD i had laying around. Getting around 250ms on the deepstack side vs 550ms using the normal windows server.
I am a little confused by the way you word this. Are you saying running DeepStack through Docker for Windows is twice as fast as running DeepStack through Docker on a Hyper-V VM?
 
Joined
Nov 6, 2019
Messages
13
Reaction score
5
Location
Spring Texas
I am still running the GentlePumpkin AITool but WOW look at these improvements!

My previous MODE=HIGH times were 300-500ms MODE=LOW times were sometimes under 200 but mostly 250ish...
Blue Iris is running on RAID0 NVME Win10Pro I7-8500 32GB RAM and I have a separate Ubuntu box running Docker/deepstack:latest with the same processor/nvme/ram:
High
high.png

Low
low.png

There seems to be no difference between HIGH/LOW so I am just going to remove the ENV entry completely until we learn about whether that is still supported. Either way, these are INSANE times, and as mentioned, will allow for practically instant triggers. I have 7 cameras that face a busyish street and now could theoretically provide about 10fps on CPU of vision detection! I cannot wait for the GPU to get finished. Each machine has a Quadro 620, which isn't super powerful, but I am happy to keep the cpu as free as possible for blue iris.
 
Last edited:

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
291
Reaction score
161
Location
UK
Hi... yes, have the latest running. But I wanted to know where to get the "log" our output of the execution time from post 2110View attachment 73117
Sorry can't help you with that question, post 2110 was made by @MnM, hopefully he can advise. It looks though it is generated from the deepstack startup window. I should add that I'm still running the original app not the forked version so it's possible the forked version has additional log data available.
 

MnM

Young grasshopper
Joined
May 14, 2014
Messages
95
Reaction score
20
Already replied here

 
Joined
Nov 6, 2019
Messages
13
Reaction score
5
Location
Spring Texas
I turned High on on mine yesterday and they differ by about 100ms. Not sure about difference to accuracy.
Some of the mistakes are comical. So far, my cat is 100% a dog, about 50% of the time. The street is sometimes an umbrella. Several mailboxes are people... LOL Not hating on this at all, it's just so random and funny
 

pbc

Getting comfortable
Joined
Jul 11, 2014
Messages
1,024
Reaction score
156
Out of the blue (seemingly anyhow) yesterday I started getting these errors which I just noticed today.

[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!
[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!
[22.10.2020, 16:33:28]: ERROR: Can't write to cameras/history.csv!

Any idea what is causing it? I'm still getting notifications.
Just bumping this.

Also noticed my processing times have gone from < 1 second many weeks ago, to 2 to 3 seconds now. Not sure if the two are related.
 

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
291
Reaction score
161
Location
UK
Just bumping this.

Also noticed my processing times have gone from < 1 second many weeks ago, to 2 to 3 seconds now. Not sure if the two are related.
I'm guessing that you have not had a response due lack of information.
It would be helpful to know software version and environment. Linux/Windows etc.
 

pbc

Getting comfortable
Joined
Jul 11, 2014
Messages
1,024
Reaction score
156
I'm guessing that you have not had a response due lack of information.
It would be helpful to know software version and environment. Linux/Windows etc.
Thanks. Running deepstack on docker on my main BI server which is a Windows 10 based server.

I'm still getting triggers to my phone so it seems to still be working. Just delayed.
 
Top