Picasso: A Modular Framework for Visualizing the Learning Process of Neural Network Image Classifiers

TL;DR

Picasso is an open-source web tool that visualizes neural network learning processes, helping researchers identify hidden issues in CNNs with customizable visualizations like occlusion and saliency maps.

Contribution

It introduces a modular framework for visualizing CNN training, supporting easy integration of new visualizations and compatibility with TensorFlow and Keras.

Findings

Supports multiple neural network architectures

Enables detection of hidden issues like proxy tasks

Facilitates minimal-configuration visualization integration

Abstract

Picasso is a free open-source (Eclipse Public License) web application written in Python for rendering standard visualizations useful for analyzing convolutional neural networks. Picasso ships with occlusion maps and saliency maps, two visualizations which help reveal issues that evaluation metrics like loss and accuracy might hide: for example, learning a proxy classification task. Picasso works with the Tensorflow deep learning framework, and Keras (when the model can be loaded into the Tensorflow backend). Picasso can be used with minimal configuration by deep learning researchers and engineers alike across various neural network architectures. Adding new visualizations is simple: the user can specify their visualization code and HTML template separately from the application code.