Testing SDK Error - Help

[Wessel Pansegrouw]

Hi There 

we are getting this error when testing - Screenshot attached.

Please let me know how to communicate with you best, I would like to buy your application but email and other forms of communication seem not to work?

Please help, I want to buy but test 1st 

Thanks so much 

[Mamadou DIOP]

Hi,

Run 'cat /proc/cpuinfo' and share the output

--
You received this message because you are subscribed to the Google Groups "doubango-ai" group.
To unsubscribe from this group and stop receiving emails from it, send an email to doubango-ai...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/c2aeee13-427a-4236-970f-90531adf44c8n%40googlegroups.com.

[Wessel Pansegrouw]

I have sent to my development team to test and give me feedback. 

Is it possible to ask a few questions while we wait for feedback:

1. I am trying to best understand your license model?

1.1 Once we buy your application are we only fixed to one device? 

1.1.1 How can we scale and roll to more devices? Do we have to pay for each unit 900Euro?

2. Is it possible to have a server in Cloud (Google / AWS) to process using the 1x license and then send results?

Hope to hear from you soon, cheers :-)

[Mamadou DIOP]

Le 6/30/2022 à 8:37 PM, Wessel Pansegrouw a écrit :

I have sent to my development team to test and give me feedback. 

Is it possible to ask a few questions while we wait for feedback:

1. I am trying to best understand your license model?

1.1 Once we buy your application are we only fixed to one device? 

1.1.1 How can we scale and roll to more devices? Do we have to pay for each unit 900Euro?

As explained at https://www.doubango.org/pricing.html, the license is per device. If you have one license you can only use it on one device/machine.

Quote from the page: "Per device: You need as many runtime licenses as devices (mobile or server). N device(s) means N runtime license(s)."

2. Is it possible to have a server in Cloud (Google / AWS) to process using the 1x license and then send results?

Yes but we only support AWS and Azure: https://github.com/DoubangoTelecom/ultimateALPR-SDK/blob/master/AWS.md

Hope to hear from you soon, cheers :-)

On Thursday, 30 June 2022 at 20:28:38 UTC+2 Mamadou wrote:

Hi,

Run 'cat /proc/cpuinfo' and share the output

Le 6/30/2022 à 8:25 PM, Wessel Pansegrouw a écrit :

Hi There 

we are getting this error when testing - Screenshot attached.

Please let me know how to communicate with you best, I would like to buy your application but email and other forms of communication seem not to work?

Please help, I want to buy but test 1st 

Thanks so much 

--
You received this message because you are subscribed to the Google Groups "doubango-ai" group.
To unsubscribe from this group and stop receiving emails from it, send an email to doubango-ai...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/c2aeee13-427a-4236-970f-90531adf44c8n%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "doubango-ai" group.
To unsubscribe from this group and stop receiving emails from it, send an email to doubango-ai...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/619307a3-2a03-4892-810d-41275aab47bfn%40googlegroups.com.

[Wessel Pansegrouw]

Hi, Thanks for the quick response 

Attached .txt file from Dev team 

[Mamadou DIOP]

You're using an old and very low-end CPU (https://ark.intel.com/content/www/us/en/ark/products/93361/intel-atom-x5z8350-processor-2m-cache-up-to-1-92-ghz.html). We use Tensorflow (https://www.tensorflow.org/) on x86 and it requires AVX instructions (https://en.wikipedia.org/wiki/Advanced_Vector_Extensions) which is 12-years old techno present on all modern CPUs.

You're trying to run the program an old low-end Atom CPU, that's not a good idea. These CPUs are not designed for servers but are equivalent to Mobiles (e.g. ARM). In such case it's better to use a mobile device with ARM CPU (faster and licenses are cheaper).

You have a quad core @1.9Ghz with 2GB memory. A Raspberry Pi (https://www.raspberrypi.com/products/raspberry-pi-4-model-b/) is faster and only cost $34.

To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/e977488e-5898-4f6a-8fe1-d07fb5a01ff5n%40googlegroups.com.

[Wessel Pansegrouw]

Ok, that makes sense. we have many other devices to test. what do you think of the spec: Intel Celeron, 4gig Ram, 32Gig HDD Linux OS?

Also, 

When we use multiple devices... The license is in the cloud(AWS), I still don't see the point of using the cloud, because the true power of your code is the fact that it's very fast and can run with no GPU and very low CPU...  Therefore, I would still need to buy a very high-end PC(Per camera) to deal with my OPEN CV app then send it to the cloud then back/ or keep it in the cloud for reporting. SO I really need to buy a license for each device - right?

Perhaps I am missing something?

Also, 

Do you also have Detection for Face (Male, Female, Age, Mood, and so on?)

Thanks for the help so far 

[Mamadou DIOP]

- "I still don't see the point of using the cloud" -> I haven't recommended using a cloud. Why are you using a cloud?

- " I would still need to buy a very high-end PC(Per camera) to deal with my OPEN CV app then send it to the cloud then back/ or keep it in the cloud for reporting" -> Haven't recommend it. The SDK can run on a mid-range x86 machine at realtime. Your CPU isn't mid-range but low-range. You'll probably have same perf issues with your "Intel Celeron". You should use a "normal"/"standard" configuration (e.g. i5, i7...). Atom and Celeron CPUs are for low power embedded devices (Mobiles)

- "the true power of your code is the fact that it's very fast and can run with no GPU and very low CPU" -> "very low CPU" maybe on ARM but on x86 it should be a descent CPU. The CPU you're trying to use is a very low-end x86 CPU even for a mobile device. As already explained, your configuration is weaker than a 35$ SBC (RPI4). Ok, the code is optimized but not for such low-end devices.

- If you have multiple cameras I'd recommend using ARM devices (e.g. Jetson). Other ANPR providers will charge you $45 per camera per month while we have a lifetime license at $39/device

To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/3e2f0559-e829-4a13-8a33-a34d21b16ed6n%40googlegroups.com.

[Wessel Pansegrouw]

I think we still miss understanding each other. 

I can't use a mobile device like android phones, The solution is required for an enterprise solution so external PCs will be needed. 

I will be required to 300+ in the first batch.

Thank you for the help. Have a good day 

[Wessel Pansegrouw]

Hi There,

When using AWS cloud processing: AWS and Azure: https://github.com/DoubangoTelecom/ultimateALPR-SDK/blob/master/AWS.md 

How many requests can we send to the server for processing? 

Example: If we have 100 cameras in different locations can we send unlimited processing requests to the cloud server all at once?

[Mamadou DIOP]

Hi,

As explained at https://www.doubango.org/pricing.html the license doesn't restrict the number of images you can process. The restriction will be defined by your bandwidth and processing power.

"Example: If we have 100 cameras in different locations can we send unlimited processing requests to the cloud server all at once?" -> No. 100 video streams requires at lot of bandwidth and processing power. You'll need multiple servers.

Sent from my Galaxy

To view this discussion on the web visit https://groups.google.com/d/msgid/doubango-ai/0164292c-5c80-413a-8942-061cd0e03999n%40googlegroups.com.

[Wessel Pansegrouw]

Ok thanks 

Would you be open to selling your source code?

Sent from my iPhone

On 05 Jul 2022, at 15:16, Mamadou DIOP <diopm...@doubango.org> wrote:



[Mamadou DIOP]

On 5 Jul 2022, at 15:48, Wessel Pansegrouw <pan...@gmail.com> wrote:

Ok thanks 

Would you be open to selling your source code?

Unfortunately no