-
Notifications
You must be signed in to change notification settings - Fork 5.5k
Custom dataset without segmentations #48
Comments
Hi @learnbott, I think it may be simpler and less error-prone to preprocess your json annotations to include the following entries:
I believe this would allow you to perform training (w/ flipping) without any modifications to the code. |
Much appreciated!
…On Sat, Jan 27, 2018 at 4:51 PM, Ilija Radosavovic ***@***.*** > wrote:
Hi @learnbott <https://github.com/learnbott>, I think it may be simpler
and less error-prone to preprocess your json annotations to include the
following entries:
'segmentation' : []
'area': box_width * box_height
'iscrowd': 0
I believe this would allow you to perform training (w/ flipping) without
any modifications to the code.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#48 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AT56PcRoCdEUiJC9NZudU7-GZy0dPVhjks5tO8R_gaJpZM4RvF6T>
.
|
@learnbott : what was the speed of wider faces detector you have trained (on HD frame) ? |
Thanks learnbott ! I don't care about CPU... |
Mine is pretty simple, but if you want a great face detector I would use Dockerface https://github.com/natanielruiz/dockerface. Its a faster r-cnn trained on multiple face datasets. https://arxiv.org/abs/1708.04370 |
@learnbott Are your errors resolved after using 'segmentation': [] As I am able to train the model after using above syntax but not able to detect objects from the validation image. Using below code for detection: cfg.MODEL.WEIGHTS = os.path.join(cfg.OUTPUT_DIR, "model_final.pth") from detectron2.utils.visualizer import ColorMode |
What about depicting bounding boxes as segmentations instead of leaving the segmentation empty? Please give me your opinion.
|
Is there a way to run a custom dataset that only has bounding boxes? I have the Wider Face dataset in COCO api json format, but it won't train without a
segmentation
field inannotations
. I had to use two work arounds:annotation['segmenation'] = 0
andannotation['area'] = bbox_height * bbox_width
(and of course leavingTRAIN.GT_MIN_AREA = -1
)TRAIN.USE_FLIPPED = True
, Detectron/lib/datasets/roidb.py had to have some code commented out in this function:With those two adjustments the code runs beautifully. Is there a flag or config param that I am missing, or an easier way to run datasets with only bboxes?
Thank you for the code and the docker version!
The text was updated successfully, but these errors were encountered: