Martlgap's picture
Update README.md
8cc31fa
|
raw
history blame
5.13 kB
metadata
license: mit
language:
  - en

Model Card for Model ID

This is a face recognition model, which extracts a facial feature vector from an aligned facial image.

This modelcard aims to be a base template for new models. It has been generated using this raw template.

Model Details

Model Description

  • Developed by: Martin Knoche
  • Funded by [optional]: Technical University of Munich
  • Shared by [optional]: Martin Knoche
  • Model type: Convolutional Neural Network
  • License: Original Work:

MIT License

Copyright (c) 2022 Zhong Yaoyao

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Changes in Code, Finetuning etc. are also under MIT License:

MIT License

Copyright (c) 2023 Martin Knoche

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Model Sources

Uses

Use the model to extract a facial feature vector from an arbitrary aligned facial image. You can then compare that vector to other facial feature vectors to decide for same or not same person.

Direct Use

The model can be used by within an ONNX-Runtime environment.

model = rt.InferenceSession("FaceTransformerOctupletLoss.onnx", providers=rt.get_available_providers())
embedding = model.run(None, {"input_image": input_image})[0][0]

input_image-Variable Dimensions: 112x112x3 Channels: Should be in RGB format Type: float Values: Between 0 and 255

embedding-Variable Dimension: 512 Type: float

Bias, Risks, and Limitations

The model was originally trained and also finetuned on the MS1M dataset. Thus please be check the MS1M dataset for bias and risks.

How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

Metrics

Accuracy [%]

Results

LFW CALFW CPLFW MLFW XQLFW
99.73 94.93 91.58 85.63 95.12

Citation

BibTeX:

@inproceedings{knoche2023octuplet,
  title={Octuplet loss: Make face recognition robust to image resolution},
  author={Knoche, Martin and Elkadeem, Mohamed and H{\"o}rmann, Stefan and Rigoll, Gerhard},
  booktitle={2023 IEEE 17th International Conference on Automatic Face and Gesture Recognition (FG)},
  pages={1--8},
  year={2023},
  organization={IEEE}
}

Model Card Author

Martin Knoche

Model Card Contact

Martin.Knoche@tum.de