anisdismail's picture
Update README.md
170ba74
|
raw
history blame
964 Bytes
metadata
language:
  - en
license: apache-2.0
tags:
  - image-classification: null
  - pytorch: null
datasets:
  - CelebA-faces
model-index:
  - name: celebA_orientation_detection_model
    results:
      - task:
          type: image_classification
          name: Image Classification
        dataset:
          type: CelebA-faces
          name: CelebA-faces
        metrics:
          - type: f1score
            value: 0.97
            name: Val F1 Score

Detecting the Orientation of CelebA pictures using Deep Learning

This model has been trained on a modified version of the CelebA dataset, which was made from flipping 20,000 and keeping 20,000 images intact. The model relies on Resnet-18 as a backbone and then is connected to one output node to classify whether images are flipped upside down (1) or not (0).