USV-based Embedded Obstacle Segmentation Submission Failed Invalid ONNX Model

[Mateus Raitz]

Hello,

I submitted my model before the deadline, but It failed the testing, the only thing it says is  "Invalid ONNX model" . I tried both ONNEX opset 11 and 13, same problem with both. My submission ID is #14227. Could I please get a more detailed information about what failed with the ONNX model. If so, would it be possible to allow me to fix it and resubmit it ?  I know the submissions should be closed by now, but I found out about this challenge just recently and spent the whole night getting my model ready, I would like to at least see how it performed in the leaderboard.  Thanks in advance. 

Best regards,

Mateus Raitz,

[Matija Tersek]

Hey,

Yes, I think we can consider your submissions. It seems that the error we are receiving is the following:

Model output shape must be [1, 1, 384, 768], got (1, 3, 48, 96)

Let me know if you'll be resubmitting the model.

Best,

Matija

Luxonis.com | DepthAI Documentation | Store

Want early access to new products? Join our free Innovation Lab to get early access to hardware and software products. You will also get the opportunity to shape core software solutions built by our development team for your use.

--
Sie erhalten diese Nachricht, weil Sie in Google Groups E-Mails von der Gruppe "MaCVi Support" abonniert haben.
Wenn Sie sich von dieser Gruppe abmelden und keine E-Mails mehr von dieser Gruppe erhalten möchten, senden Sie eine E-Mail an macvi-suppor...@googlegroups.com.
Diese Diskussion finden Sie unter https://groups.google.com/d/msgid/macvi-support/3cb22fd1-ae3c-4afe-aa54-d022eeee5160n%40googlegroups.com.
Weitere Optionen finden Sie unter https://groups.google.com/d/optout.

[Mateus Raitz]

Should it not be [1, 3, 384, 768] instead? You wrote  [1,1,384,768]. But I get what you meant, I checked in Netron now and the input is fine but the output is 1,3,48,96 for some reason. I would fix it and would like to re-submit it. 

Thanks!

Mateus Raitz. 

[Mateus Raitz]

I have fixed the issue and resubmit it , my new ID # is 14239. It is being processed by the test pipeline now. Thanks! 

Best regards,

Mateus Raitz.

[Mateus Raitz]

Sorry that was the validation , the real submission ID # is 14240.

Best regards,

Mateus Raitz.

[Mateus Raitz]

The validation failed again, thus the real submission will also fails, so I will just cancel that one. Could you please check why validation submission  #14239 failed?

Best regards,

Mateus Raitz.

[Mateus Raitz]

I think it maybe failed because the challenge requires the tensor output to be argmaxed, I try adding that and will validate, if it pass I will submit it to the challenge. 

Best regards,

Mateus Raitz.

[Matija Tersek]

Yes, output has to be argmaxed. 

Best,

Matija

Luxonis.com | DepthAI Documentation | Store

Want early access to new products? Join our free Innovation Lab to get early access to hardware and software products. You will also get the opportunity to shape core software solutions built by our development team for your use.

--
Sie erhalten diese Nachricht, weil Sie in Google Groups E-Mails von der Gruppe "MaCVi Support" abonniert haben.
Wenn Sie sich von dieser Gruppe abmelden und keine E-Mails mehr von dieser Gruppe erhalten möchten, senden Sie eine E-Mail an macvi-suppor...@googlegroups.com.

Diese Diskussion finden Sie unter https://groups.google.com/d/msgid/macvi-support/1ee66c37-db9f-4eb8-8c90-6193dc99e9e9n%40googlegroups.com.

[Mateus Raitz]

It passed the validation now, thanks so much for your help. I  just submitted it to the challenge. It is ID  #14242.

Best regards,

Mateus Raitz.

[Mateus Raitz]

It passed validation, but when submitting it now got "Script execution failed. Make sure the format of your submission is correct". Would it be possible to get more information about what is wrong with the format of my submission?

Best regards,

Mateus Raitz.

[Benjamin Kiefer]

@Matija upload has been deactivated for these challenge tracks. Pls rename the script and reactivate uploading to allow this specific upload :) 

[Mateus Raitz]

Please let me know when the upload is activated again, I am not being able to upload it anymore.

Best regards,

Mateus Raitz.

[ZQ Z]

Hello, organizer. The competition has ended. Can I still upload onnx now? When will the final ranking results be announced?

Mateus Raitz 在 2024年12月21日 星期六清晨6:42:36 [UTC+8] 的信中寫道：

[Matija Tersek]

> Please let me know when the upload is activated again, I am not being able to upload it anymore.

Will try to parse results for the submitted ONNX directly and let you know when we have the results.

Best,

Matija

Luxonis.com | DepthAI Documentation | Store

Want early access to new products? Join our free Innovation Lab to get early access to hardware and software products. You will also get the opportunity to shape core software solutions built by our development team for your use.

Diese Diskussion finden Sie unter https://groups.google.com/d/msgid/macvi-support/cb9d71ba-d94a-4be3-a54c-88db473075d0n%40googlegroups.com.

[Mateus Raitz]

Sounds good, thansk! Will my submission  be in the Leaderboard ?

Best regards,

Mateus Raitz.

[Matija Tersek]

Hi Mateus,

It's visible as "Resubmission - 14242", but unfortunately, it wasn't fast enough.

Best,

Matija

Luxonis.com | DepthAI Documentation | Store

Want early access to new products? Join our free Innovation Lab to get early access to hardware and software products. You will also get the opportunity to shape core software solutions built by our development team for your use.

Diese Diskussion finden Sie unter https://groups.google.com/d/msgid/macvi-support/2b2f3766-d3bd-42c8-bbd5-00ab7951ed3fn%40googlegroups.com.