Is a neural network better than Ash at detecting Team Rocket? If so, how?


Our whole existence is a never-ending riddle. Are we the only ones in the Universe? What's the point of life? Is a neural network better than Ash at recognizing Team Rocket? The first two are non-trivial questions that keep many scientists and philosophers up at night. The last one, however, keeps me up at night. In this project, I'll attempt to answer it.

This wep application uses TensorFlow.js to run an object detection model trained in Google Cloud AutoML and an image classifier trained in TensorFlow with the purpose of producing and visualizating the network's activation maps.

Select an image

Prediction output:


Object Detection

Activations Maps

conv2d layer activations

Filter number:

max_pooling2d activations

Filter number:

conv2d_1 activations

Filter number:

max_pooling2d_1 activations

Filter number:

conv2d_2 activations

Filter number:

max_pooling2d_2 activations

Filter number: