Jump to content
  • 0

custom model to detect object and save/share information


Ajander Singh
 Share

Question

Hi,

I am new to AMS so please bear with me.

I am in requirement to detect object using custom TensorFlow model from live stream and store detected objects name and frame information in my application db.

1. How can I use my own trained model for object detection.
2. is it possible to draw bounding box for detected model in live stream so viewer can get highlighted object.
3. How can call some API to send/share detected object and frame information our API. Is it possible using webhooks in AMS?

Please suggest on above.

Thanks
Link to comment
Share on other sites

  • Answers 4
  • Created
  • Last Reply

Top Posters For This Question

4 answers to this question

Recommended Posts

  • 0
Hi Ajander Singh,
Unfortunately our tensorflow object detection api is very simple and does not have the features that you want. It only detects objects based on a fixed model creates some screenshots where bounding box of rectangle of detected objects are drawn in it. Then you can get which objects are detected and the probabilities for them using our REST API method:
1-) currently you cannot do that.
2-) That is currently also not possible
3) that is not implemented yet but I can add a feature request for this if you want.

I can add feature requests for these features if you want which can be implemented in the next versions.

Best Regards,
Can

PS: Free or Paid one-on-one support and consultancy for your streaming project, please book at https://calendly.com/antmedia/

On Sat, 13 Jun at 8:43 PM , Ajander Singh <ajander.singh@gmail.com> wrote:
Hi,

I am new to AMS so please bear with me.

I am in requirement to detect object using custom TensorFlow model from live stream and store detected objects name and frame information in my application db.

1. How can I use my own trained model for object detection.
2. is it possible to draw bounding box for detected model in live stream so viewer can get highlighted object.
3. How can call some API to send/share detected object and frame information our API. Is it possible using webhooks in AMS?

Please suggest on above.

Thanks
--
You received this message because you are subscribed to the Google Groups "Ant Media Server" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ant-media-server+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ant-media-server/481a72f8-c062-4539-9e8b-15dc95ccacc7o%40googlegroups.com.
9035:1264838
Link to comment
Share on other sites

  • 0
Yes, Please add as feature request.

Thanks

On Monday, 15 June 2020 01:16:43 UTC+5:30, Ant Media wrote:
Hi Ajander Singh,
Unfortunately our tensorflow object detection api is very simple and does not have the features that you want. It only detects objects based on a fixed model creates some screenshots where bounding box of rectangle of detected objects are drawn in it. Then you can get which objects are detected and the probabilities for them using our REST API method:
1-) currently you cannot do that.
2-) That is currently also not possible
3) that is not implemented yet but I can add a feature request for this if you want.

I can add feature requests for these features if you want which can be implemented in the next versions.

Best Regards,
Can

PS: Free or Paid one-on-one support and consultancy for your streaming project, please book at https://calendly.com/antmedia/

On Sat, 13 Jun at 8:43 PM
, Ajander Singh <
> wrote:
Hi,

I am new to AMS so please bear with me.

I am in requirement to detect object using custom TensorFlow model from live stream and store detected objects name and frame information in my application db.

1. How can I use my own trained model for object detection.
2. is it possible to draw bounding box for detected model in live stream so viewer can get highlighted object.
3. How can call some API to send/share detected object and frame information our API. Is it possible using webhooks in AMS?

Please suggest on above.

Thanks
--

You received this message because you are subscribed to the Google Groups "Ant Media Server" group.

To unsubscribe from this group and stop receiving emails from it, send an email to
.

To view this discussion on the web visit
.

9035:1264838
Link to comment
Share on other sites

  • 0
Hi Ajander Singh,

no, unfortunately we dont have plugin support but I talked with the team if you buy a perpetual or yearly license, we can give you full source code of enterprise edition and we can point you where the tensorflow object detection is implemented. Then I think you can add all kind of machine learning models and draw on the frames bounding boxes.
But beware we can only point you to the classes, we cannot give you any detailed explanation because it will take take from our senior developers. You need understand how everything works by looking to the code.

If you want full source code, please buy a yearly or perpetual license and write me. I will send you the full source code.

Best Regards,
Can



PS: Free or Paid one-on-one support and consultancy for your streaming project, please book at https://calendly.com/antmedia/

On Mon, 15 Jun at 8:04 PM , Ajander Singh <ajander.singh@gmail.com> wrote:
Hi,

Can I develop my own plugins for media server or if I can access media pipeline to capture the stream/frame to hook my own code?

Thanks

--
You received this message because you are subscribed to the Google Groups "Ant Media Server" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ant-media-server+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ant-media-server/1d348e17-aec7-413f-8f45-84ffbbddf73fo%40googlegroups.com.

9035:1264838
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Answer this question...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share


×
×
  • Create New...