GAN Lab: Understanding Complex Deep Generative Models using Interactive Visual Experimentation

Abstract

Recent success in deep learning has generated immense interest among practitioners and students, inspiring many to learn about this new technology. While visual and interactive approaches have been successfully developed to help people more easily learn deep learning, most existing tools focus on simpler models. In this work, we present GAN Lab, the first interactive visualization tool designed for non-experts to learn and experiment with Generative Adversarial Networks (GANs), a popular class of complex deep learning models. With GAN Lab, users can interactively train generative models and visualize the dynamic training process's intermediate results. GAN Lab tightly integrates an model overview graph that summarizes GAN's structure, and a layered distributions view that helps users interpret the interplay between submodels. GAN Lab introduces new interactive experimentation features for learning complex deep learning models, such as step-by-step training at multiple levels of abstraction for understanding intricate training dynamics. Implemented using TensorFlow.js, GAN Lab is accessible to anyone via modern web browsers, without the need for installation or specialized hardware, overcoming a major practical challenge in deploying interactive tools for deep learning.

arXiv commentThis paper will be published in the IEEE Transactions on Visualization and Computer Graphics, 25(1), January 2019, and presented at IEEE VAST 2018

References in corpus (8)

• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Generative Adversarial Networks: An Overview
• Understanding Neural Networks Through Deep Visualization
• NIPS 2016 Tutorial: Generative Adversarial Networks
• A note on the evaluation of generative models
• Unrolled Generative Adversarial Networks
• Do Convolutional Neural Networks Learn Class Hierarchy?
• Direct-Manipulation Visualization of Deep Networks

Cited by in corpus (36)

• A Survey of the Recent Architectures of Deep Convolutional Neural Networks
• CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualization
• A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
• explAIner: A Visual Analytics Framework for Interactive and Explainable Machine Learning
• The State of the Art in Enhancing Trust in Machine Learning Models with the Use of Visualizations
• TensorFlow.js: Machine Learning for the Web and Beyond
• Review: Deep Learning in Electron Microscopy
• Explaining Vulnerabilities to Adversarial Machine Learning through Visual Analytics
• Net2Vis -- A Visual Grammar for Automatically Generating Publication-Tailored CNN Architecture Visualizations
• ConfusionFlow: A model-agnostic visualization for temporal analysis of classifier confusion
• medigan: a Python library of pretrained generative models for medical image synthesis
• Diagnosing Concept Drift with Visual Analytics
• A Survey on Neural Network Interpretability
• Model Compression in Practice: Lessons Learned from Practitioners Creating On-device Machine Learning Experiences
• CNN 101: Interactive Visual Learning for Convolutional Neural Networks
• NNVA: Neural Network Assisted Visual Analysis of Yeast Cell Polarization Simulation
• AIive: Interactive Visualization and Sonification of Neural Networks in Virtual Reality
• GAM Coach: Towards Interactive and User-centered Algorithmic Recourse
• A Survey of Visual Analytics Techniques for Machine Learning
• AHMoSe: A Knowledge-Based Visual Support System for Selecting Regression Machine Learning Models
• Should We Trust (X)AI? Design Dimensions for Structured Experimental Evaluations
• MARLens: Understanding Multi-agent Reinforcement Learning for Traffic Signal Control via Visual Analytics
• CNNPruner: Pruning Convolutional Neural Networks with Visual Analytics
• Pattern Detection on Glioblastoma's Waddington landscape via Generative Adversarial Networks
• exploRNN: Understanding Recurrent Neural Networks through Visual Exploration
• HypoML: Visual Analysis for Hypothesis-based Evaluation of Machine Learning Models
• Visual Evaluation of Generative Adversarial Networks for Time Series Data
• WebSHAP: Towards Explaining Any Machine Learning Models Anywhere
• Level generation and style enhancement -- deep learning for game development overview
• Towards Better Modeling with Missing Data: A Contrastive Learning-based Visual Analytics Perspective
• A Game-Based Approach for Helping Designers Learn Machine Learning Concepts
• A Visual Analytics Framework for Explaining and Diagnosing Transfer Learning Processes
• Transformer Explainer: Learning LLM Transformers with Interactive Visual Explanation and Experimentation
• Time-Series Regeneration with Convolutional Recurrent Generative Adversarial Network for Remaining Useful Life Estimation
• Generative Deep Learning for Virtuosic Classical Music: Generative Adversarial Networks as Renowned Composers
• Interactive Visual Study of Multiple Attributes Learning Model of X-Ray Scattering Images