Hello, what is the meaning of the -annotations parameter?

[gzchenjiajun]

Hello, I have a question: --annotations parameter is used, I really did not understand, now I filled in the parameters of my prediction picture, but he asked me to fill in x1, x2, y1, Y2, but this is my predictive picture, I don't know what it means to fill this. I have a guess, is it true that if the prediction succeeds, the bounding box's coordinate values will overwrite the contents of --annotations?

[eg4000]

Hi, Please follow the instructions from the Usage section in the README.md file. We did not use the --annotations argument.

[gzchenjiajun]

Hi, Please follow the instructions from the Usage section in the README.md file. We did not use the --annotations argument.

I know, but the code USES this -- annotation, and it passes in an argument to execute, and in fact, I have to put in a valid CSV file and a path to execute, so I don't really understand how an annotation works

[eg4000]

It has a default value

[gzchenjiajun]

It has a default value

Annotation it has a default value: data_dir + '/annotations_test.csv' If I'm not mistaken, this is test_data, which is the predict image I passed in, so the question goes back to what I said above, why am I supposed to fill in x1,y1,x2,y2, this prediction image I don't have x1,y1,x2,y2. Or is it just like my guess, after successful prediction, the CSV will be overwritten by predict result? These are the predicted bounding boxes Sorry, I have a few more questions. Thank you very much for answering

[eg4000]

I'm not sure I understand your question. You should have an _annotationstest.csv file.

[gzchenjiajun]

I'm not sure I understand your question. You should have an _annotationstest.csv file. hi, my question here is why do I write x1,x2,y1,y2 here in annotations_test.csv file? So this is my prediction picture, and I don't know what x1,x2,y1,y2 are there

[gzchenjiajun]

I'm not sure I understand your question. You should have an _annotationstest.csv file.

hi, my question here is why do I write x1,x2,y1,y2 here in annotations_test.csv file? So this is my prediction picture, and I don't know what x1,x2,y1,y2 are there

[gzchenjiajun]

I'm not sure I understand your question. You should have an _annotationstest.csv file.

Hi, I would like to ask whether I took the h5 pre-training weight you provided, meaning that I went straight to step3, namely predict, and now I took iou_resnet50_csv_06. For the weight of h5, I went straight to step3, was this an error?

[eg4000]

The prediction results are saved in CSV format in the "results" folder and drawn in "res_images_iou" folder. The test ground-truth annotations are not required for prediction but are available if you want to evaluate the results. See #5.

[gzchenjiajun]

The prediction results are saved in CSV format in the "results" folder and drawn in "res_images_iou" folder. The test ground-truth annotations are not required for prediction but are available if you want to evaluate the results. See #5.

The folder of "res_images_iou" is empty, the CSV file in "results" has no data, I think I may have made a mistake, I will try a new image, do you have any Suggestions for the use of images?

[gzchenjiajun]

I'll try again. Thank you

[gzchenjiajun]

The prediction results are saved in CSV format in the "results" folder and drawn in "res_images_iou" folder. The test ground-truth annotations are not required for prediction but are available if you want to evaluate the results. See #5.

The function filter_duplicate_candidates returns Empty DataFrame, I am wondering if there is something wrong with the test picture I entered (but I threw in clear sku picture), I will try another picture, thank you

[gzchenjiajun]

The prediction results are saved in CSV format in the "results" folder and drawn in "res_images_iou" folder. The test ground-truth annotations are not required for prediction but are available if you want to evaluate the results. See #5.

Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 7257 MB memory) -> physical GPU (device: 0, name: Quadro P4000, pci bus id: 0000:01:00.0, compute capability: 6.1) Loading images... 100%|█████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 5675.65it/s] Loading model, this may take a second... /home/ubuntu/anaconda3/lib/python3.6/site-packages/keras/engine/saving.py:292: UserWarning: No training configuration found in save file: the model was not compiled. Compile it manually. warnings.warn('No training configuration found in save file: ' Created folder /home/ubuntu/Desktop/share1/SKU110K_CVPR19-master/res_images_iou Traceback (most recent call last): File "object_detector_retinanet/keras_retinanet/bin/predict.py", line 155, in main() File "object_detector_retinanet/keras_retinanet/bin/predict.py", line 150, in main hard_score_rate=hard_score_rate File "/home/ubuntu/Desktop/share1/SKU110K_CVPR19-master/object_detector_retinanet/keras_retinanet/utils/predict_iou.py", line 113, in predict writer.writerows(csv_data_lst) TypeError: a bytes-like object is required, not 'str'

This is the error prediction I conducted after loading the pre-training. I wonder if the weight was not loaded into the keras model. Leading to an inability to predict correctly, which leads to the problems above? Can you help me? I think I'm getting the answer. Thank you very, very much.

[gzchenjiajun]

The prediction results are saved in CSV format in the "results" folder and drawn in "res_images_iou" folder. The test ground-truth annotations are not required for prediction but are available if you want to evaluate the results. See #5.

My friend, I have checked and found that my version meets the requirements of readme.md. I think I will be running soon. Can you help me.thanks

[gzchenjiajun]

I'm running. Thank you for your help.

[4vik]

I'm running. Thank you for your help.

Hi, I am running into the same problem as you (No training configuration found in saved pre-trained model file). How did you end up fixing it? I'd appreciate your help a lot, got a deadline coming up... @gzchenjiajun