Blue Iris and CodeProject.AI ALPR

[wpiman]

I just got one of these for my Dual M2 TPU Coral adapter...

Dual Edge TPU Adapter - PCIe x1 Low Profile

Dual Edge TPU Adapter is designed for Coral m.2 Accelerator with Dual Edge TPU to be used on a system with PCIe x1 slot available.

www.makerfabs.com

Currently, I am having BI use Codeproject on a Cent OS box running docker. I also have a couple of Orange PI 5s that run Codeproject and use the RKNN accelerator.

Does anyone else have one of these? What machine are you installing it on? I might put it in my Windows Box (Xeon) that BI runs on... I don't like the windows install for Codeproject (PITA) but maybe that is the way to go...

 

[hopalong]

UPDATE

Hmm maybe that last re-install got it to work, see benchmarks below, though should I worry about all the warnings I see listed in the logs below?








__ORIGINAL_
Anyone else getting this System Log error message with the ALPR module and 2.2.4-Beta?

10:59:16:Started License Plate Reader module
10:59:16:ALPR_adapter.py: Traceback (most recent call last):
10:59:16:ALPR_adapter.py:   File "/app/modules/ALPR/ALPR_adapter.py", line 8, in
10:59:16:ALPR_adapter.py:     from module_runner import ModuleRunner
10:59:16:ALPR_adapter.py:   File "/app/modules/ALPR/../../SDK/Python/module_runner.py", line 28, in
10:59:16:ALPR_adapter.py:     import aiohttp
10:59:16:ALPR_adapter.py: ModuleNotFoundError: No module named 'aiohttp'
10:59:16:Module ALPR has shutdown

I've tried stopping all modules. Uninstalling and reinstalling the ALPR module a few times. Here are the install error messages I am seeing:

11:15:27:Current Version is 2.2.4-Beta 11:15:42:Preparing to install module 'ALPR' 11:15:42:Downloading module 'ALPR' 11:15:42:Installing module 'ALPR' 11:15:42:Installer script at '/app/setup.sh' 11:15:42:ALPR: Hi Docker! We will disable shared python installs for downloaded modules 11:15:42:ALPR: No schemas installed 11:15:42:ALPR: (No schemas means: we can't detect if you're in light or dark mode) 11:15:42:ALPR: sh: 1: lsmod: not found 11:15:42:ALPR: Installing CodeProject.AI Analysis Module 11:15:42:ALPR: ====================================================================== 11:15:42:ALPR: CodeProject.AI Installer 11:15:42:ALPR: ====================================================================== 11:15:42:ALPR: 20.03 GiB available 11:15:42:ALPR: Checking GPU support 11:15:42:ALPR: CUDA Present...No 11:15:42:ALPR: Allowing GPU Support: Yes 11:15:42:ALPR: Allowing CUDA Support: Yes 11:15:42:ALPR: General CodeProject.AI setup 11:15:42:ALPR: Setting permissions...Done 11:15:42:ALPR: Processing SDK 11:15:50:ALPR: Adding toolchain repo... Done 11:15:52:ALPR: Checking for installed dependencies (from gcc-4.9 apt-utils)...Done 11:15:55:ALPR: W: https://packages.cloud.google.com/apt/dists/coral-edgetpu-stable/InRelease: Key is stored in legacy trusted.gpg keyring (/etc/apt/trusted.gpg), see the DEPRECATION section in apt-key(8) for details. 11:15:56:ALPR: Installing missing dependencies: gcc-4.9...Done 11:15:58:ALPR: Upgrading libstdc... Done 11:16:05:ALPR: Checking for installed dependencies (from ca-certificates gnupg libc6-dev libfontconfig1 libgdiplus libjpeg-dev zlib1g-dev)...Done 11:16:05:ALPR: All dependencies already installed. 11:16:13:ALPR: Checking for installed dependencies (from ffmpeg libsm6 libxext6 mesa-utils curl rsync unzip wget)...Done 11:16:13:ALPR: All dependencies already installed. 11:16:13:ALPR: The command could not be loaded, possibly because: 11:16:13:ALPR: * You intended to execute a .NET application: 11:16:13:ALPR: The application '--version' does not exist. 11:16:13:ALPR: * You intended to execute a .NET SDK command: 11:16:13:ALPR: No .NET SDKs were found. 11:16:13:ALPR: Download a .NET SDK: 11:16:13:ALPR: https://aka.ms/dotnet/download 11:16:13:ALPR: Learn about SDK resolution: 11:16:13:ALPR: Checking for .NET >= 7.0...All good. Current .NET is , requested was 7.0 11:16:13:ALPR: https://aka.ms/dotnet/sdk-not-found 11:16:13:ALPR: Installing module ALPR 2.6 11:16:13:ALPR: Python 3.8 is already installed 11:16:14:ALPR: Installing Virtual Environment tools for Linux... Done 11:16:33:ALPR: Creating Virtual Environment (Local)... Done 11:16:33:ALPR: Checking for Python 3.8...Found Python 3.8.18. present 11:16:33:ALPR: Checking for CUDA...Not found 11:16:34:ALPR: Installing PIP... Done 11:16:36:ALPR: Ensuring PIP compatibility... Done 11:17:18:ALPR: Updating PIP... Done 11:17:21:ALPR: Installing setuptools... Done 11:17:21:ALPR: Choosing packages from requirements.linux.txt 11:21:07:ALPR: Installing Packages into Virtual Environment... Success 11:21:13:ALPR: Downloading OCR models...Expanding... Done. 11:21:13:ALPR: Applying PaddleOCR patch 11:21:13:ALPR: Installing Server SDK support: 11:21:13:ALPR: Checking for CUDA...Not found 11:21:14:ALPR: Installing PIP... Done 11:21:18:ALPR: Ensuring PIP compatibility... Done 11:21:24:ALPR: Updating PIP... Done 11:21:30:ALPR: Installing setuptools... Done 11:21:30:ALPR: Choosing packages from requirements.txt 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3 already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3-2.0.5.dist-info already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/PIL already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow.libs already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow-9.5.0.dist-info already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna already exists. Specify --upgrade to force replacement. 11:21:41:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna-3.4.dist-info already exists. Specify --upgrade to force replacement. 11:21:42:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/charset_normalizer already exists. Specify --upgrade to force replacement. 11:21:42:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi already exists. Specify --upgrade to force replacement. 11:21:42:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi-2023.7.22.dist-info already exists. Specify --upgrade to force replacement. 11:21:42:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests already exists. Specify --upgrade to force replacement. 11:21:42:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests-2.31.0.dist-info already exists. Specify --upgrade to force replacement. 11:21:43:ALPR: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/bin already exists. Specify --upgrade to force replacement. 11:21:43:Call to Install on module ALPR has completed. 11:21:44:ALPR: Installing Packages into Virtual Environment... Success 11:21:44:ALPR: Setup complete 11:21:44:Module ALPR installed successfully. 11:21:44:Installer exited with code 0 11:21:44: 11:21:44: 11:21:44:Attempting to start ALPR with /app/modules/ALPR/bin/linux/python38/venv/bin/python3 "/app/modules/ALPR/ALPR_adapter.py" 11:21:44:Module 'License Plate Reader' 2.6 (ID: ALPR) 11:21:44:Module Path: /app/modules/ALPR 11:21:44:AutoStart: True 11:21:44:Queue: alpr_queue 11:21:44:Platforms: windows,linux,macos,macos-arm64 11:21:44:GPU: Support disabled 11:21:44:Parallelism: 0 11:21:44:Accelerator: 11:21:44:Half Precis.: enable 11:21:44:Runtime: python38 11:21:44:Runtime Loc: Local 11:21:44:FilePath: ALPR_adapter.py 11:21:44:Pre installed: False 11:21:44:Start pause: 1 sec 11:21:44:LogVerbosity: 11:21:44:Valid: True 11:21:44:Environment Variables 11:21:44:AUTO_PLATE_ROTATE = True 11:21:44:OCR_OPTIMAL_CHARACTER_HEIGHT = 60 11:21:44:OCR_OPTIMAL_CHARACTER_WIDTH = 36 11:21:44:OCR_OPTIMIZATION = True 11:21:44:PLATE_CONFIDENCE = 0.7 11:21:44:PLATE_RESCALE_FACTOR = 2 11:21:44:PLATE_ROTATE_DEG = 0 11:21:44: 11:21:44:Started License Plate Reader module 11:21:45:Module ALPR started successfully. 11:22:04:ALPR_adapter.py: CPAI_MODULE_REQUIRED_MB not found. Setting to default 0

Here's a screenshot with all the red warnings above:

 

[wpiman]

I bought this for my dual Coral TPU...

Dual Edge TPU Adapter - PCIe x1 Low Profile

Dual Edge TPU Adapter is designed for Coral m.2 Accelerator with Dual Edge TPU to be used on a system with PCIe x1 slot available.

www.makerfabs.com

Adding my dual M2 card, plugged it into my Windows box (ugh- don't like windows)-- then deleted Codeproject and did a fresh install...

The object detection times appear to be 4 times faster than a similar Linux machine with the same processor for a photo I chose.... Don't think that the LPR makes use of the TPUs.. (faster time is TPU obviously)






I then put the latest image on my Orange PI 5- and installed the RKNN object detection and disable the others. That is by far the fastest object detection in my setup.... Has to be the cheapest as well...

 

[colipto]

Yep something weird going on with alpr and GPUs... could be my specific setup but based on previous comments that helped me get my working i would say its the latter.

My specific setup

VMware ESXI i7-12700k

• VM 1 - Blue Iris Server Windows 2022, latest blue iris update
• VM2 - Code Project Server Windows 2022 passthrough GTX 1050, latest 12.2 Cuda drivers
• 
• 

What i have noticed is that i couldn't get ALPR detection to work with GPU... if license plate reader is set to GPU with my setup i do not get any OCR / license plate reads. Could not get 6.2 working with GPU(CUDA) correctly..that may be in part due to my old gpu.. revertring to v5 .Net resolved my issue on that part.

I also tried this on my desktop with a 2080 TI and experienced the same issue

Edit: thank you to all who have posted in here and have helped me track down this issue.. now i can focus on getting my night time detection 100%

 

[MikeLud1]

Yep something weird going on with alpr and GPUs... could be my specific setup but based on previous comments that helped me get my working i would say its the latter.

My specific setup

VMware ESXI i7-12700k

• VM 1 - Blue Iris Server Windows 2022, latest blue iris update
• VM2 - Code Project Server Windows 2022 passthrough GTX 1050, latest 12.2 Cuda drivers
• View attachment 173230
• View attachment 173231

What i have noticed is that i couldn't get ALPR detection to work with GPU... if license plate reader is set to GPU with my setup i do not get any OCR / license plate reads. Could not get 6.2 working with GPU(CUDA) correctly..that may be in part due to my old gpu.. revertring to v5 .Net resolved my issue on that part.

I also tried this on my desktop with a 2080 TI and experienced the same issue

Edit: thank you to all who have posted in here and have helped me track down this issue.. now i can focus on getting my night time detection 100%

There is two issues for the GTX 1050 ALPR v2.6 we are having issues the some of the Python packages and need the find the right combination for it to work.
The second issues is you have CUDA 12.2 install. The ALPR module does not work with CUDA 12.x it only works with CUDA 11.x.
To fix the CUDA version you need to uninstall CUDA 12.2 and install CUDA 11.8 and cuDNN v8.9.5 for CUDA 11.x. If you downgrade to CUDA 11.8 your 2080 TI should work

 

[colipto]

There is two issues for the GTX 1050 ALPR v2.6 we are having issues the some of the Python packages and need the find the right combination for it to work.
The second issues is you have CUDA 12.2 install. The ALPR module does not work with CUDA 12.x it only works with CUDA 11.x.
To fix the CUDA version you need to uninstall CUDA 12.2 and install CUDA 11.8 and cuDNN v8.9.5 for CUDA 11.x. If you downgrade to CUDA 11.8 your 2080 TI should work

After downgrading to 11.8 and installing cuDNN, several restarts and more importantly using the test image provided by codeproject it does seem to work on my 2080 TI machine..

I'm still having issues with the GTX 1050 which will be the primary codeproject server .. what can i do to resolve this so i can use my GPU for ALPR? The only way to get plate reads is to switch to CPU on ALPR module

 

[MikeLud1]

After downgrading to 11.8 and installing cuDNN, several restarts and more importantly using the test image provided by codeproject it does seem to work on my 2080 TI machine..

I'm still having issues with the GTX 1050 which will be the primary codeproject server .. what can i do to resolve this so i can use my GPU for ALPR? The only way to get plate reads is to switch to CPU on ALPR module
View attachment 173310

Are you getting any errors in the log, if please post them.
I have a GTX 1050 Ti that I could try running some tests this weekend so I can see what is happening.

 

[colipto]

No errors in log from what i can see

The below results with CPU(detecting and working) then i switch ALPR over to GPU (No plates found)


01:43:11:Response received (...f8281f): Found Plate: DL11479
01:43:12:Client request 'alpr' in queue 'alpr_queue' (...5c466c)
01:43:12:Request 'alpr' dequeued from 'alpr_queue' (...5c466c)
01:43:12:License Plate Reader: Retrieved alpr_queue command
01:43:12:Request 'custom' dequeued from 'objectdetection_queue' (...419a9a)
01:43:12:Client request 'custom' in queue 'objectdetection_queue' (...419a9a)
01:43:12:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:43:12:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:43:12:Response received (...419a9a): Found DayPlate
01:43:12:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...419a9a) took 36ms
01:43:12:Response received (...5c466c): Found Plate: DL11479
01:43:12:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...5c466c) took 175ms
01:43:12:Request 'alpr' dequeued from 'alpr_queue' (...2b923a)
01:43:12:Client request 'alpr' in queue 'alpr_queue' (...2b923a)
01:43:12:License Plate Reader: Retrieved alpr_queue command
01:43:12:Client request 'custom' in queue 'objectdetection_queue' (...1fc021)
01:43:12:Request 'custom' dequeued from 'objectdetection_queue' (...1fc021)
01:43:12:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:43:12:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:43:12:Response received (...1fc021): Found DayPlate
01:43:12:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...1fc021) took 67ms
01:43:13:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...2b923a) took 309ms
01:43:13:Response received (...2b923a): Found Plate: DL 11479
01:43:13:Request 'alpr' dequeued from 'alpr_queue' (...3b890c)
01:43:13:Client request 'alpr' in queue 'alpr_queue' (...3b890c)
01:43:13:License Plate Reader: Retrieved alpr_queue command
01:43:13:Client request 'custom' in queue 'objectdetection_queue' (...fb286e)
01:43:13:Request 'custom' dequeued from 'objectdetection_queue' (...fb286e)
01:43:13:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:43:13:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:43:13:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...fb286e) took 59ms
01:43:13:Response received (...fb286e): Found DayPlate
01:43:13:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...3b890c) took 180ms
01:43:13:Response received (...3b890c): Found Plate: DL 11479
01:44:01:Sending shutdown request to python/ALPR
01:44:01:Client request 'Quit' in queue 'alpr_queue' (...a070dd)
01:44:01:Request 'Quit' dequeued from 'alpr_queue' (...a070dd)
01:44:01:License Plate Reader: Retrieved alpr_queue command
01:44:11:Module ALPR has shutdown
01:44:11:ALPR_adapter.py: has exited
01:44:34:ALPR went quietly
01:44:34:GetCommandByRuntime: Runtime=python37, Location=Local
01:44:34:
01:44:34:Module 'License Plate Reader' 2.6 (ID: ALPR)
01:44:34:Module Path: C:\Program Files\CodeProject\AI\modules\ALPR
01:44:34:AutoStart: True
01:44:34:Queue: alpr_queue
01:44:34latforms: windows,linux,macos,macos-arm64
01:44:34:Command: C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python37\venv\scripts\Python
01:44:34:Starting C:\Program Files...ws\python37\venv\scripts\Python "C:\Program Files...\modules\ALPR\ALPR_adapter.py"
01:44:34:
01:44:34:Attempting to start ALPR with C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python37\venv\scripts\Python "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py"
01:44:34:GPU: Support enabled
01:44:34arallelism: 0
01:44:34:Accelerator:
01:44:34:Half Precis.: enable
01:44:34:Runtime: python37
01:44:34:Runtime Loc: Local
01:44:34:FilePath: ALPR_adapter.py
01:44:34re installed: False
01:44:34:Start pause: 1 sec
01:44:34:LogVerbosity:
01:44:34:Valid: True
01:44:34:Environment Variables
01:44:34:AUTO_PLATE_ROTATE = True
01:44:34:OCR_OPTIMAL_CHARACTER_HEIGHT = 60
01:44:34:OCR_OPTIMAL_CHARACTER_WIDTH = 36
01:44:34:OCR_OPTIMIZATION = True
01:44:34LATE_CONFIDENCE = 0.7
01:44:34LATE_RESCALE_FACTOR = 2
01:44:34LATE_ROTATE_DEG = 0
01:44:34:
01:44:34:Started License Plate Reader module
01:44:41:ALPR_adapter.py: Timeout connecting to the server
01:44:41:ALPR_adapter.py: CPAI_MODULE_REQUIRED_MB not found. Setting to default 0
01:44:41:ALPR_adapter.py: License Plate Reader started.License Plate Reader: License Plate Reader started.
01:44:41:License Plate Reader: License Plate Reader started.
01:45:05:Client request 'alpr' in queue 'alpr_queue' (...f9f53d)
01:45:05:Request 'alpr' dequeued from 'alpr_queue' (...f9f53d)
01:45:05:License Plate Reader: Retrieved alpr_queue command
01:45:05:Client request 'custom' in queue 'objectdetection_queue' (...45af0f)
01:45:05:Request 'custom' dequeued from 'objectdetection_queue' (...45af0f)
01:45:05:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:45:05:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:45:05:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...45af0f) took 72ms
01:45:05:Response received (...45af0f): Found DayPlate
01:45:06:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...f9f53d) took 927ms
01:45:06:Response received (...f9f53d): No plates found
01:45:16:Client request 'alpr' in queue 'alpr_queue' (...71110b)
01:45:16:Request 'alpr' dequeued from 'alpr_queue' (...71110b)
01:45:16:License Plate Reader: Retrieved alpr_queue command
01:45:16:Client request 'custom' in queue 'objectdetection_queue' (...a49f5b)
01:45:16:Request 'custom' dequeued from 'objectdetection_queue' (...a49f5b)
01:45:16:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:45:17:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:45:17:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...a49f5b) took 77ms
01:45:17:Response received (...a49f5b): Found DayPlate
01:45:17:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...71110b) took 121ms
01:45:17:Response received (...71110b): No plates found
01:45:17:Client request 'alpr' in queue 'alpr_queue' (...028d8f)
01:45:17:Request 'alpr' dequeued from 'alpr_queue' (...028d8f)
01:45:17:License Plate Reader: Retrieved alpr_queue command
01:45:17:Request 'custom' dequeued from 'objectdetection_queue' (...95db2f)
01:45:17:Client request 'custom' in queue 'objectdetection_queue' (...95db2f)
01:45:17:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command
01:45:17:Object Detection (YOLOv5 6.2): Detecting using license-plate
01:45:18:Object Detection (YOLOv5 6.2): Rec'd request for Object Detection (YOLOv5 6.2) command 'custom' (...95db2f) took 91ms
01:45:18:Response received (...95db2f): Found DayPlate
01:45:18:License Plate Reader: Rec'd request for License Plate Reader command 'alpr' (...028d8f) took 164ms
01:45:18:Response received (...028d8f): No plates found

 

[colipto]

Did notice this error related to yolo on the same machine.. clearly having issues with GPU aswell so i have to use .net - i did verify cudnn was installed

01:59:03:Object Detection (YOLOv5 6.2): [RuntimeError] : Traceback (most recent call last):
File "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo\detect.py", line 141, in do_detection
det = detector(img, size=640)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 705, in forward
y = self.model(x, augment=augment) # forward
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 515, in forward
y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 209, in forward
return self._forward_once(x, profile, visualize) # single-scale inference, train
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 121, in _forward_once
x = m(x) # run
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 245, in forward
return self.conv(torch.cat((x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]), 1))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 57, in forward
return self.act(self.bn(self.conv(x)))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\conv.py", line 463, in forward
return self._conv_forward(input, self.weight, self.bias)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\conv.py", line 460, in _conv_forward
self.padding, self.dilation, self.groups)
RuntimeError: cuDNN error: CUDNN_STATUS_NOT_INITIALIZED

 

[colipto]

Okay.. simply downloading the bat script is not installing cudnn... i realize that after reviewing the installation guide here Installation Guide.. although i didnt have to install cudnn fully on my 2080 ti setup... still dealing with issues on 1050

now dealing with the following errors in logs

02:07:54:Object Detection (YOLOv5 6.2): [OutOfMemoryError] : Traceback (most recent call last):
File "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo\detect.py", line 141, in do_detection
det = detector(img, size=640)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 705, in forward
y = self.model(x, augment=augment) # forward
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 515, in forward
y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 209, in forward
return self._forward_once(x, profile, visualize) # single-scale inference, train
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 121, in _forward_once
x = m(x) # run
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 140, in forward
return self.cv4(self.act(self.bn(torch.cat((y1, y2), 1))))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 57, in forward
return self.act(self.bn(self.conv(x)))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\batchnorm.py", line 182, in forward
self.eps,
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\functional.py", line 2451, in batch_norm
input, weight, bias, running_mean, running_var, training, momentum, eps, torch.backends.cudnn.enabled
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 2.00 MiB (GPU 0; 2.00 GiB total capacity; 294.85 MiB already allocated; 0 bytes free; 328.00 MiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF


and finally

02:12:10:Object Detection (YOLOv5 6.2): [RuntimeError] : Traceback (most recent call last):
File "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo\detect.py", line 141, in do_detection
det = detector(img, size=640)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 705, in forward
y = self.model(x, augment=augment) # forward
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 515, in forward
y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 209, in forward
return self._forward_once(x, profile, visualize) # single-scale inference, train
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 121, in _forward_once
x = m(x) # run
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 245, in forward
return self.conv(torch.cat((x[..., ::2, ::2], x[..., 1::2, ::2], x[..., ::2, 1::2], x[..., 1::2, 1::2]), 1))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\common.py", line 57, in forward
return self.act(self.bn(self.conv(x)))
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\conv.py", line 463, in forward
return self._conv_forward(input, self.weight, self.bias)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\conv.py", line 460, in _conv_forward
self.padding, self.dilation, self.groups)
RuntimeError: cuDNN error: CUDNN_STATUS_ALLOC_FAILED

have to stick to using .net Yolo v5 and cpu alpr for now

 

[105437]

Anyone do the in-place upgrade to v2.3 on a Windows PC running Blue Iris?

 

[MikeLud1]

Anyone do the in-place upgrade to v2.3 on a Windows PC running Blue Iris?

I am on v2.3

 

[105437]

I am on v2.3

I'm thinking of installing it over the top of v2.1.8 and see how it does.

 

[MikeLud1]

I'm thinking of installing it over the top of v2.1.8 and see how it does.

I installed it over v2.2.4.

 

[105437]

I installed it over v2.2.4.

I have a P400, is YOLO v5.3.1 still the better performing (speed/accuracy) choice over v5.6.2?

 

[MikeLud1]

With your GPU v5 6.2 should work fine with CUDA 11.8. The only issue would have is the 2GB VRAM.

 

[105437]

With your GPU v5 6.2 should work fine with CUDA 11.8. The only issue would have is the 2GB VRAM.

I didn't upgrade CUDA, it's currently v10.1. Should I upgrade?

 

[MikeLud1]

I didn't upgrade CUDA, it's currently v10.1. Should I upgrade?

If it working well for you on v5 3.1 maybe say on this version

 

[105437]

If it working well for you on v5 3.1 maybe say on this version

I'm trying v5.6.2 as a test. Seems to be running okay so far. Just curious if upgrading CUDA would help performance.

 

[xm4rcell0x]

Hi all! i'm trying to use ALPR module with my Tesla P4, but i can't. I'm on Unraid and my CPAI docker is on it. I can't even choose to enable or disable the GPU fpr this module.
Here there are some screenshot



thank you so much