/Identify-Specific-Objects-in-an-Image-or-Video-using-machine-learning-in-the-browser

This application uses Tensorflow.js to serve a mobilenet SSD image classification model in the browser which is used to crop (using cropper.js) and identify specific objects in an image or a video.

Primary LanguageJavaScriptMIT LicenseMIT

Identify-Specific-Objects-in-an-Image-or-Video-using-machine-learning-in-the-browser

This projects aims to crop specific parts of an Image and video and run an image classification model using tensorflow.js on the cropped Image in the browser

Demo


demo video source: https://www.youtube.com/watch?v=Yi_kDT5L5s4

References


https://fengyuanchen.github.io/cropperjs/
https://www.tensorflow.org/js/tutorials/transfer/image_classification
https://developer.mozilla.org/en-US/docs/Web/API/Canvas_API/Manipulating_video_using_canvas

License

MIT