Open xyzdcgan opened 5 years ago
Hello @xyzdcgan do you have any solution for your problem?
I have same problem.
Dear xyzdcgan ,
I want to use Mask-RCNN to train on my custom dataset (just one class).
First of all, I use Labelme to create groundtruth for each image. Each image has a .json file as its ground-truth.
After that, I don't know what should I do...
I want to know how to convert my groundtruth file to a format that Mask-rcnn can use. Do you have any tools ?
Looking forward to hearing from you, thanks in advance.
With best regards, Ming
Hi @fevzidas @hjm1990818
steps i followed to use mask-rcnn:
1) Labelled each custom images with polygon 2) Save that files in json format 3) Used a pretrained model 4) Trained that model with my images
It was working fine
https://github.com/priya-dwivedi/Deep-Learning/tree/master/mask_rcnn_damage_detection
I used above link for all things to detect using mask-rcnn
Hi @fevzidas @hjm1990818
steps i followed to use mask-rcnn:
- Labelled each custom images with polygon
Dear xyzdcgan,
Thank you very much for your response.
Maybe you don't understand what I mean, I am not a English native.
Could you please give me your e-mail address? I want to show some examples for you for a detailed explaination.
Looking forward to hearing from you, thanks.
With best regards, MIng
priyanksolanki900@gmail.com
I am creating custom mask-rcnn object detection model. I labeled the data with polygon shape and created csv file.
I am not able to create tf record from csv as, it has polygon set rather then rectangle bounding box. Can you help in creating tf record..?
Also i am using this tensroflow, and using mask_rcnn_resnet101_atrous_coco from https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md
And configuring config file and giving path for it.
https://tensorflow-object-detection-api-tutorial.readthedocs.io/en/latest/training.html
Using this kind of implementation for mask-rcnn, so I can get checkpoint file. Rather then using a pre trained h5 model of mask-rcnn
Thanks