{ "id": 416691, "title": "Could not check whether submission is allowed.", "url": "/competitions/hubmap-hacking-the-human-vasculature/discussion/416691", "author_name": "", "post_date": "2023-06-12T16:21:30.369120800Z", "votes": 3, "comment_count": 11, "views": 0, "content": "

Hello all! 😁

\n

First of all, I'm really happy to start my journey in Kaggle with some friends, but we are newcomers and we are facing two unexpected errors.

\n\n
submission_dict = {\n    : [],\n    : [],\n    : [],\n    : [],\n}\n\n img  images:\n\n    results = model(img, conf=, verbose=)\n    res = results[]\n\n     (res.boxes) == :\n        \n\n    image_id = res.path.split()[-].split()[]\n\n    confidences = res.boxes.conf.tolist()\n    masks = res.masks.data.cpu().numpy()\n    masks = [np.array(cv2.resize(mask, (, ))).astype()  mask  masks]\n    encoded_masks = [encode_binary_mask(m)  m  masks]\n    encoded_strings = [\n                     conf, mask  (confidences, encoded_masks)]\n    final_string = .join(encoded_strings)\n\n    submission_dict[].append(image_id)\n    submission_dict[].append()    \n    submission_dict[].append()    \n    submission_dict[].append(final_string) \n\nsubmission_df = pd.DataFrame(submission_dict)\nsubmission_df = submission_df.set_index()\nsubmission_df.to_csv()\n
\n\n

What can we do with both errors?

\n

All help is highly appreciated. Thank you so much. 🤗

\n

Cheers.

", "messages": [ { "id": "2297506", "postDate": "06/12/2023 16:21:30", "content": "

Hello all! 😁

\n

First of all, I'm really happy to start my journey in Kaggle with some friends, but we are newcomers and we are facing two unexpected errors.

\n\n
submission_dict = {\n    : [],\n    : [],\n    : [],\n    : [],\n}\n\n img  images:\n\n    results = model(img, conf=, verbose=)\n    res = results[]\n\n     (res.boxes) == :\n        \n\n    image_id = res.path.split()[-].split()[]\n\n    confidences = res.boxes.conf.tolist()\n    masks = res.masks.data.cpu().numpy()\n    masks = [np.array(cv2.resize(mask, (, ))).astype()  mask  masks]\n    encoded_masks = [encode_binary_mask(m)  m  masks]\n    encoded_strings = [\n                     conf, mask  (confidences, encoded_masks)]\n    final_string = .join(encoded_strings)\n\n    submission_dict[].append(image_id)\n    submission_dict[].append()    \n    submission_dict[].append()    \n    submission_dict[].append(final_string) \n\nsubmission_df = pd.DataFrame(submission_dict)\nsubmission_df = submission_df.set_index()\nsubmission_df.to_csv()\n
\n\n

What can we do with both errors?

\n

All help is highly appreciated. Thank you so much. 🤗

\n

Cheers.

", "rawMarkdown": "Hello all! 😁\n\nFirst of all, I'm really happy to start my journey in Kaggle with some friends, but we are newcomers and we are facing two unexpected errors.\n\n- We have tried to submit several times within 2 days and we are not able to make a submission with our YOLOv8m-seg baseline. This is our code for generating CSV and we are getting ***Submission Scoring Error***.\n\n```python\nsubmission_dict = {\n \"id\": [],\n \"height\": [],\n \"width\": [],\n \"prediction_string\": [],\n}\n\nfor img in images:\n\n results = model(img, conf=0.4, verbose=False)\n res = results[0]\n\n if len(res.boxes) == 0:\n continue\n \n image_id = res.path.split('/')[-1].split('.')[0]\n\n confidences = res.boxes.conf.tolist()\n masks = res.masks.data.cpu().numpy()\n masks = [np.array(cv2.resize(mask, (512, 512))).astype(bool) for mask in masks]\n encoded_masks = [encode_binary_mask(m) for m in masks]\n encoded_strings = [f\"0 {conf:.5f} {mask.decode('utf-8')}\"\n for conf, mask in zip(confidences, encoded_masks)]\n final_string = \" \".join(encoded_strings)\n\n submission_dict[\"id\"].append(image_id)\n submission_dict[\"height\"].append(512) \n submission_dict[\"width\"].append(512) \n submission_dict[\"prediction_string\"].append(final_string) \n\nsubmission_df = pd.DataFrame(submission_dict)\nsubmission_df = submission_df.set_index('id')\nsubmission_df.to_csv('/kaggle/working/submission.csv')\n```\n\n- Right now, I am not able to do any submission as Kaggle tells me: ***Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.***\n\nWhat can we do with both errors?\n\nAll help is highly appreciated. Thank you so much. 🤗\n\nCheers.", "votes": null }, { "id": "2297565", "postDate": "06/12/2023 16:55:41", "content": "

Yes, i found this problem and I don't know the reason why \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.\"

", "rawMarkdown": "Yes, i found this problem and I don't know the reason why \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.\"", "votes": null }, { "id": "2297567", "postDate": "06/12/2023 16:56:42", "content": "

Hey there, good luck with the competition.

\n

We're a part of the Image Matching Challenge hosted on kaggle and have encountered the same error \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered\" in the last hour and a half .
\nI guess it has something to do with the servers.

", "rawMarkdown": "Hey there, good luck with the competition.\n\nWe're a part of the Image Matching Challenge hosted on kaggle and have encountered the same error \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered\" in the last hour and a half .\nI guess it has something to do with the servers.", "votes": null }, { "id": "2297651", "postDate": "06/12/2023 17:50:11", "content": "

Thank you so much for your feedback. It seems that the servers have gone down for a couple of hours. But now, it seems OK!

", "rawMarkdown": "Thank you so much for your feedback. It seems that the servers have gone down for a couple of hours. But now, it seems OK!", "votes": null }, { "id": "2297653", "postDate": "06/12/2023 17:50:53", "content": "

Thank you so much for your feedback. For me now it is okay. Good luck with the competition!

", "rawMarkdown": "Thank you so much for your feedback. For me now it is okay. Good luck with the competition!", "votes": null }, { "id": "2300143", "postDate": "06/13/2023 03:53:11", "content": "

I am facing the same issue with the YOLO V8 model. Did you managed to fix this?

", "rawMarkdown": "I am facing the same issue with the YOLO V8 model. Did you managed to fix this?", "votes": null }, { "id": "2300370", "postDate": "06/13/2023 07:00:57", "content": "

I'm sorry but we are still at the same point. No solution has been reached yet.

", "rawMarkdown": "I'm sorry but we are still at the same point. No solution has been reached yet.", "votes": null }, { "id": "2300395", "postDate": "06/13/2023 07:21:15", "content": "

Very strange problem indeed, tried + 10 attempts all failures. Interestingly my submission which uses UNET based model worked fine earlier. Since I have used the same submission format i don't see why a submission format using YOLO-v8 gives an error. Is there something wrong with how YOLOV8 predicts the masks? I will update if I manage to resolve this. For now it might be useful to move to other models. : (

", "rawMarkdown": "Very strange problem indeed, tried + 10 attempts all failures. Interestingly my submission which uses UNET based model worked fine earlier. Since I have used the same submission format i don't see why a submission format using YOLO-v8 gives an error. Is there something wrong with how YOLOV8 predicts the masks? I will update if I manage to resolve this. For now it might be useful to move to other models. : (", "votes": null }, { "id": "2300466", "postDate": "06/13/2023 07:57:28", "content": "

That's an interesting fact…

\n

I mean YOLOv8-seg returns a res.masks.data [NxHxW, where N is the number of objects]. Then I resize it to (512, 512) since I work in (640, 640) and then convert it to boolean. That's my input to the encode_binary_mask… I don't know what or where the failure can be.

\n

I hope we both can find solution soon. :)

", "rawMarkdown": "That's an interesting fact...\n\nI mean YOLOv8-seg returns a res.masks.data [NxHxW, where N is the number of objects]. Then I resize it to (512, 512) since I work in (640, 640) and then convert it to boolean. That's my input to the encode_binary_mask... I don't know what or where the failure can be.\n\nI hope we both can find solution soon. :)", "votes": null }, { "id": "2300480", "postDate": "06/13/2023 08:09:22", "content": "

Okay it worked!!

\n

So I added \"\" empty strings for images with no predictions and it worked! <y lb score is 0.181 now. Here is what i did.

\n
    \n
  1. Replaced all predictions by \"\" in my UNET notebook = score 0
    2. \n
  2. Replace all predictions by \"\" for all images in test in YOLO notebook = score 0
    3. \n
  3. Replace all predicitons by \"\" when prediction mask is empty (checked using result.masks.data) = score 0.181
    4. \n
\n

I hope it works for you too. Seems all test image id needs to be present in the submission csv for the scoring to work properly. Here is the code snippet

\n

``
\nOutputDict ={}

\n

allResults = list(result)

\n

listTest = os.listdir(TEST_PATH)
\nlistTest = [os.path.join(TEST_PATH,x) for x in listTest]
\nfor fil in tqdm(listTest):
\n result = model.predict(fil, imgsz=512,conf = 0.25, verbose=False)
\n for r in result:
\n pathImage = r.path
\n iden = os.path.split(pathImage)[-1].split(\".\")[0]
\n boxes = r.boxes.xywh
\n confidence = r.boxes.conf
\n if r.masks:
\n masks = r.masks.data
\n predString= convYOLOtoCode(masks, confidence)
\n OutputDict[iden] = predString
\n else:
\n OutputDict[iden] = \"\"
\n``

", "rawMarkdown": "Okay it worked!! \n\nSo I added \"\" empty strings for images with no predictions and it worked! \"\"

\n

Here is the code snippet

", "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3951798%2Fbaa1c9c1b66a14dadec66c47274d86a4%2FCapture.PNG?generation=1686643929155819&alt=media)\n\n\nHere is the code snippet", "votes": null }, { "id": "2300812", "postDate": "06/13/2023 13:15:54", "content": "

Yes! It worked! It is a matter of submitting empty strings (\"\") in images without detections!

\n

Thank you so much my friend 😁

", "rawMarkdown": "Yes! It worked! It is a matter of submitting empty strings (\"\") in images without detections!\n\nThank you so much my friend 😁", "votes": null } ], "comments": [ { "id": 2297565, "author_name": "hidngnguyna", "author_url": "", "post_date": "06/12/2023 16:55:41", "content": "

Yes, i found this problem and I don't know the reason why \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.\"

", "votes": null, "replies": [ { "id": 2297653, "author_name": "santimontieleu", "author_url": "", "post_date": "06/12/2023 17:50:53", "content": "

Thank you so much for your feedback. For me now it is okay. Good luck with the competition!

", "votes": null, "replies": [] } ] }, { "id": 2297567, "author_name": "jovankamchev", "author_url": "", "post_date": "06/12/2023 16:56:42", "content": "

Hey there, good luck with the competition.

\n

We're a part of the Image Matching Challenge hosted on kaggle and have encountered the same error \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered\" in the last hour and a half .
\nI guess it has something to do with the servers.

", "votes": null, "replies": [ { "id": 2297651, "author_name": "santimontieleu", "author_url": "", "post_date": "06/12/2023 17:50:11", "content": "

Thank you so much for your feedback. It seems that the servers have gone down for a couple of hours. But now, it seems OK!

", "votes": null, "replies": [] } ] }, { "id": 2300143, "author_name": "dhruvkhatri", "author_url": "", "post_date": "06/13/2023 03:53:11", "content": "

I am facing the same issue with the YOLO V8 model. Did you managed to fix this?

", "votes": null, "replies": [ { "id": 2300370, "author_name": "santimontieleu", "author_url": "", "post_date": "06/13/2023 07:00:57", "content": "

I'm sorry but we are still at the same point. No solution has been reached yet.

", "votes": null, "replies": [ { "id": 2300395, "author_name": "dhruvkhatri", "author_url": "", "post_date": "06/13/2023 07:21:15", "content": "

Very strange problem indeed, tried + 10 attempts all failures. Interestingly my submission which uses UNET based model worked fine earlier. Since I have used the same submission format i don't see why a submission format using YOLO-v8 gives an error. Is there something wrong with how YOLOV8 predicts the masks? I will update if I manage to resolve this. For now it might be useful to move to other models. : (

", "votes": null, "replies": [ { "id": 2300466, "author_name": "santimontieleu", "author_url": "", "post_date": "06/13/2023 07:57:28", "content": "

That's an interesting fact…

\n

I mean YOLOv8-seg returns a res.masks.data [NxHxW, where N is the number of objects]. Then I resize it to (512, 512) since I work in (640, 640) and then convert it to boolean. That's my input to the encode_binary_mask… I don't know what or where the failure can be.

\n

I hope we both can find solution soon. :)

", "votes": null, "replies": [ { "id": 2300480, "author_name": "dhruvkhatri", "author_url": "", "post_date": "06/13/2023 08:09:22", "content": "

Okay it worked!!

\n

So I added \"\" empty strings for images with no predictions and it worked! <y lb score is 0.181 now. Here is what i did.

\n
    \n
  1. Replaced all predictions by \"\" in my UNET notebook = score 0
    2. \n
  2. Replace all predictions by \"\" for all images in test in YOLO notebook = score 0
    3. \n
  3. Replace all predicitons by \"\" when prediction mask is empty (checked using result.masks.data) = score 0.181
    4. \n
\n

I hope it works for you too. Seems all test image id needs to be present in the submission csv for the scoring to work properly. Here is the code snippet

\n

``
\nOutputDict ={}

\n

allResults = list(result)

\n

listTest = os.listdir(TEST_PATH)
\nlistTest = [os.path.join(TEST_PATH,x) for x in listTest]
\nfor fil in tqdm(listTest):
\n result = model.predict(fil, imgsz=512,conf = 0.25, verbose=False)
\n for r in result:
\n pathImage = r.path
\n iden = os.path.split(pathImage)[-1].split(\".\")[0]
\n boxes = r.boxes.xywh
\n confidence = r.boxes.conf
\n if r.masks:
\n masks = r.masks.data
\n predString= convYOLOtoCode(masks, confidence)
\n OutputDict[iden] = predString
\n else:
\n OutputDict[iden] = \"\"
\n``

", "votes": null, "replies": [ { "id": 2300486, "author_name": "dhruvkhatri", "author_url": "", "post_date": "06/13/2023 08:12:34", "content": "

\"\"

\n

Here is the code snippet

", "votes": null, "replies": [ { "id": 2300812, "author_name": "santimontieleu", "author_url": "", "post_date": "06/13/2023 13:15:54", "content": "

Yes! It worked! It is a matter of submitting empty strings (\"\") in images without detections!

\n

Thank you so much my friend 😁

", "votes": null, "replies": [] } ] } ] } ] } ] } ] } ] } ], "raw_markdown_by_id": { "2297506": "Hello all! 😁\n\nFirst of all, I'm really happy to start my journey in Kaggle with some friends, but we are newcomers and we are facing two unexpected errors.\n\n- We have tried to submit several times within 2 days and we are not able to make a submission with our YOLOv8m-seg baseline. This is our code for generating CSV and we are getting ***Submission Scoring Error***.\n\n```python\nsubmission_dict = {\n \"id\": [],\n \"height\": [],\n \"width\": [],\n \"prediction_string\": [],\n}\n\nfor img in images:\n\n results = model(img, conf=0.4, verbose=False)\n res = results[0]\n\n if len(res.boxes) == 0:\n continue\n \n image_id = res.path.split('/')[-1].split('.')[0]\n\n confidences = res.boxes.conf.tolist()\n masks = res.masks.data.cpu().numpy()\n masks = [np.array(cv2.resize(mask, (512, 512))).astype(bool) for mask in masks]\n encoded_masks = [encode_binary_mask(m) for m in masks]\n encoded_strings = [f\"0 {conf:.5f} {mask.decode('utf-8')}\"\n for conf, mask in zip(confidences, encoded_masks)]\n final_string = \" \".join(encoded_strings)\n\n submission_dict[\"id\"].append(image_id)\n submission_dict[\"height\"].append(512) \n submission_dict[\"width\"].append(512) \n submission_dict[\"prediction_string\"].append(final_string) \n\nsubmission_df = pd.DataFrame(submission_dict)\nsubmission_df = submission_df.set_index('id')\nsubmission_df.to_csv('/kaggle/working/submission.csv')\n```\n\n- Right now, I am not able to do any submission as Kaggle tells me: ***Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.***\n\nWhat can we do with both errors?\n\nAll help is highly appreciated. Thank you so much. 🤗\n\nCheers.", "2297565": "Yes, i found this problem and I don't know the reason why \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered.\"", "2297567": "Hey there, good luck with the competition.\n\nWe're a part of the Image Matching Challenge hosted on kaggle and have encountered the same error \"Could not check whether submission is allowed. Please try again. System Account CompetitionsSubmitValidator is not registered\" in the last hour and a half .\nI guess it has something to do with the servers.", "2297651": "Thank you so much for your feedback. It seems that the servers have gone down for a couple of hours. But now, it seems OK!", "2297653": "Thank you so much for your feedback. For me now it is okay. Good luck with the competition!", "2300143": "I am facing the same issue with the YOLO V8 model. Did you managed to fix this?", "2300370": "I'm sorry but we are still at the same point. No solution has been reached yet.", "2300395": "Very strange problem indeed, tried + 10 attempts all failures. Interestingly my submission which uses UNET based model worked fine earlier. Since I have used the same submission format i don't see why a submission format using YOLO-v8 gives an error. Is there something wrong with how YOLOV8 predicts the masks? I will update if I manage to resolve this. For now it might be useful to move to other models. : (", "2300466": "That's an interesting fact...\n\nI mean YOLOv8-seg returns a res.masks.data [NxHxW, where N is the number of objects]. Then I resize it to (512, 512) since I work in (640, 640) and then convert it to boolean. That's my input to the encode_binary_mask... I don't know what or where the failure can be.\n\nI hope we both can find solution soon. :)", "2300480": "Okay it worked!! \n\nSo I added \"\" empty strings for images with no predictions and it worked!