tf records generation from labels.npz #30
Conversation
…l-maker into tf_object_detection
|
@Geoyi reading through and trying the demo now. Is there a way to write |
|
theoretically yes @drewbo, but I only worked with dataframe/csv so far. |
drewbo
left a comment
drewbo
left a comment
There was a problem hiding this comment.
Noted a few small changes inline. I'll come through afterwards and add the folder/text changes and fix some typos
label_maker/label.py
Outdated
| cl_features = len([bb for l in labels for bb in l if bb[4] == i]) | ||
| cl_tiles = len([l for l in labels if len(list(filter(_bbox_class(i), l)))]) # pylint: disable=cell-var-from-loop | ||
| cl_features = len([bb for l in labels for bb in l if bb[4] == i + 1]) | ||
| cl_tiles = len([l for l in labels if len(list(filter(_bbox_class(i+1), l)))]) # pylint: disable=cell-var-from-loop |
There was a problem hiding this comment.
@Geoyi can you remove this commit? This change was already made on master so it's causing the merge conflict
There was a problem hiding this comment.
I thought I have resolved it, will doable check.
| @@ -0,0 +1,153 @@ | |||
| # Example Use: A building object detection with TensorFlow API | |||
|
|
|||
| You have to walk through all these steps to be able to train a TensorFlow Object Detection and have the following results: | |||
There was a problem hiding this comment.
@Geoyi Can you add a short overview of what TensorFlow Object Detection API is and why we are using it?
There was a problem hiding this comment.
an overview has added to the head of the walkthrough too.
| for bbox in bboxes: | ||
| if bbox[4] == 1: | ||
| cl_str = "building" | ||
| if bbox[0] < 0: |
showcase using label maker with Tensorflow object detection particularly.
@drewbo, can you review this code and document? I wanna make sure this code is pythonic enough, and if you could test it out that would be great too.