https://github.com/lucianopalmeida/coreml-mnist
A simple CoreML handwritten digits classification demo
https://github.com/lucianopalmeida/coreml-mnist
coreml coreml-mnist digit ios-app prediction swift4 xcode9
Last synced: 6 months ago
JSON representation
A simple CoreML handwritten digits classification demo
- Host: GitHub
- URL: https://github.com/lucianopalmeida/coreml-mnist
- Owner: LucianoPAlmeida
- Created: 2018-06-28T07:13:17.000Z (almost 8 years ago)
- Default Branch: master
- Last Pushed: 2019-01-06T23:42:08.000Z (over 7 years ago)
- Last Synced: 2024-12-26T13:13:33.499Z (over 1 year ago)
- Topics: coreml, coreml-mnist, digit, ios-app, prediction, swift4, xcode9
- Language: Swift
- Homepage:
- Size: 22.6 MB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# CoreML-MNIST
[](https://opensource.org/licenses/MIT)
### CoreML is Awesome \o/
This is a simple Demo App to learn the CoreML API.
This app make digits predictions based on a model trained with the MNIST Dataset of handwritten digits.
It just draw the digit the CoreML will make a prediction.
### The Model
The model was created and trained using [Tensorflow](https://www.tensorflow.org/) and Converted to a CoreML model using
[tf-coreml](https://github.com/tf-coreml/tf-coreml).
The network code, the conversion code and the models generated are available on [LucianoPAlmeida/mnist-number-classification](https://github.com/LucianoPAlmeida/mnist-number-classification).
### Credits and Thanks
- [Brian Advent](https://www.youtube.com/channel/UCysEngjfeIYapEER9K8aikw) for the awesome [Core ML Tutorial: Create a Simple Machine Learning App - Image Classification](https://www.youtube.com/watch?v=NNKPbdT9gXU) video.
- [CoreML Tools](https://github.com/apple/coremltools) and [tf-coreml](https://github.com/tf-coreml/tf-coreml) to make possible the conversion of my tensorflow model to a CoreML model.
## Licence
CoreML-MNIST is released under the [MIT License](https://opensource.org/licenses/MIT).