Loading
Cookies help us deliver our services. By using our services, you agree to our use of cookies. Learn more

crowdAI is shutting down - please read our blog post for more information

Mapping Challenge

Building Missing Maps with Machine Learning


Completed
719
Submissions
1074
Participants
49292
Views

Error during submission.

Posted by sarath almost 2 years ago

raise CrowdAIExecuteFunctionError(payload[“message”]) crowdai.challenges.crowdai_errors.CrowdAIExecuteFunctionError: Argument ‘bb’ has incorrect type (expected numpy.ndarray, got list)

I get this error during evaluation. Data type of my annotation keys are same as that in Random Submission.ipynb, I’ve cross-checked it multiple times. Here is a sample from my predictions, {‘score’: 0.99, ‘image_id’: 36659, ‘category_id’: 100, ‘segmentation’: [[238, 1, 238, 6, 242, 10, 246, 20, 254, 25, 261, 36, 272, 36, 283, 25, 283, 14, 275, 1]], ‘bbox’: [238, 1, 45, 35]}

Posted by Ursus  almost 2 years ago |  Quote

I had the same, you need to make a submission at rle format like here https://github.com/crowdAI/crowdai-mapping-challenge-mask-rcnn/blob/master/Prediction-and-Submission.ipynb _mask = maskUtils.encode(np.asfortranarray(mask)) _mask["counts"] = _mask["counts"].decode("UTF-8") _result["segmentation"] = _mask

Posted by sarath  almost 2 years ago |  Quote

Hi @Ursus

I noticed that segmentation is in rle format in Prediction-and-Submission.ipynb. But server seems to accept poly format too. I had no errors with submission made from “Random Submission.ipynb” which uses poly format.

Posted by Ursus  almost 2 years ago |  Quote

Hi, @sarath I saw your submission was graded, how did you solve the problem?

Posted by ChePeter  almost 2 years ago |  Quote

Hi, @sarath I saw your submission was graded, how did you solve the problem?

Posted by ChePeter  almost 2 years ago |  Quote

Hi, @sarath I saw your submission was graded, how did you solve the problem?

Posted by sarath  almost 2 years ago |  Quote

@Ursus and @ChePeter

I don’t have a clear idea why the problem occurred. It was probably because my model didn’t predict any buildings for few images. After adding random annotations for these images my submission got accepted.