---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: mask
    dtype: image
  splits:
  - name: train
    num_bytes: 119307954.15
    num_examples: 2450
  - name: validation
    num_bytes: 27229918.0
    num_examples: 613
  - name: test
    num_bytes: 39479824.0
    num_examples: 955
  download_size: 169673709
  dataset_size: 186017696.15
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---

# Links
* Paper: https://arxiv.org/pdf/1908.09101v2
* Repository: https://github.com/Mhaiyang/ICCV2019_MirrorNet
* Project page: https://mhaiyang.github.io/ICCV2019_MirrorNet/index.html
* We got our data from: https://github.com/Charmve/Mirror-Glass-Detection

# Split info
We split the train to train and validation with the ratio 80% and 20% respectively. If you want to use the original split, you can just combine train and validation.

# License info
Refer to the project page, original repository, and paper. We retrieve the dataset from third party repository.