zhiqic / chartreader Goto Github PK

Hey @zhiqic 😀 - I was wondering if you could share a .json file of bounding box test predictions for the public EC400K test set (e.g. in the COCO format: [x,y,width,height]). This way you don't have to share the trained model weights but we would still be able to run our metrics and have your best model as a baseline in our research paper. This would be super helpful and we'd be happy to acknowledge your contribution!

Best, Till

the trained baseline files' link is unuseful [https://drive.google.com/file/d/1qtCLlzKm8mx7kQOV1criUbqcGnNh58Rr/view?usp=sharing]

Hello,

Could you share with me the model .pkl file? I would like to evaluate it for my research thesis.

Thanks in advance,
Navina

Hi. has the repo been moved? Will it be re-posted?

I passed the training data set, it raises index error. The categories_id in dataset and code are not matched. Can you please specify clearer the id numbers? I got index errors and try to fix it according to the training dataset, which doesn't work and made things way more complicated.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 6 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 5 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 6 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 4 is out of bounds for axis 1 with size 3
0%| | 4/50000 [00:00<27:28, 30.34it/s]Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 6 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 5 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 4 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 5 is out of bounds for axis 1 with size 3
0%| | 8/50000 [00:00<26:41, 31.21it/s]Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 4 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 6 is out of bounds for axis 1 with size 3
Shuffling...
Data extraction error occurred.
Traceback (most recent call last):
File "/home/ubuntu/ChartReader/train_extraction.py", line 127, in train
training, ind = sample_data(training_db, ind)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ubuntu/ChartReader/sampling_function.py", line 528, in sample_data
center_heatmaps[b_ind, category, yce, xce] = 1
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 4 is out of bounds for axis 1 with size 3
Too many extraction errors. Terminating...
0%|

I followed the readme file and trained models for step 1 and 2, but didn't seem to know how to get the Component Position & Type Prediction model or output.

I have a question about the KPDetection and the KPGrouping. Do they use the same dataset to train? Why is the argument data_dir different? The training dataset is only one. But there are a lot of errors erupt coming out while I am training the KPGrouping. More then that, the result of the KPDetection is not well.

The top picture is the KPDetection result using the default config from json file KPDetection.json which is in the config directory.

我进行了训练detection的部分，在bar chart上得到了很好的结果，但在line chart和bar chart上得到的结果很差，并注意到train部分的代码是“bar only”，请问我讲如何更改得以训练识别line chart和pie chart呢

Where is the pretrained model? Is there have some code tell me how to use model to inference on some dataset? Or is there have a huggingFace line to use your model?