CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualization

Abstract

Deep learning's great success motivates many practitioners and students to learn about this exciting technology. However, it is often challenging for beginners to take their first step due to the complexity of understanding and applying deep learning. We present CNN Explainer, an interactive visualization tool designed for non-experts to learn and examine convolutional neural networks (CNNs), a foundational deep learning model architecture. Our tool addresses key challenges that novices face while learning about CNNs, which we identify from interviews with instructors and a survey with past students. CNN Explainer tightly integrates a model overview that summarizes a CNN's structure, and on-demand, dynamic visual explanation views that help users understand the underlying components of CNNs. Through smooth transitions across levels of abstraction, our tool enables users to inspect the interplay between low-level mathematical operations and high-level model structures. A qualitative user study shows that CNN Explainer helps users more easily understand the inner workings of CNNs, and is engaging and enjoyable to use. We also derive design lessons from our study. Developed using modern web technologies, CNN Explainer runs locally in users' web browsers without the need for installation or specialized hardware, broadening the public's education access to modern deep learning techniques.

11 pages, 14 figures, to be presented at IEEE VIS 2020. For a demo video, see https://youtu.be/HnWIHWFbuUQ . For a live demo, visit https://poloclub.github.io/cnn-explainer/

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• TensorFlow: A system for large-scale machine learning
• Understanding Neural Networks Through Deep Visualization
• GAN Lab: Understanding Complex Deep Generative Models using Interactive Visual Experimentation
• Do Convolutional Neural Networks Learn Class Hierarchy?
• TensorFlow.js: Machine Learning for the Web and Beyond
• Direct-Manipulation Visualization of Deep Networks
• CNN 101: Interactive Visual Learning for Convolutional Neural Networks
• Analyzing the Noise Robustness of Deep Neural Networks

Cited by in corpus (20)

• Review: Deep Learning in Electron Microscopy
• Damage detection using in-domain and cross-domain transfer learning
• Augmented Math: Authoring AR-Based Explorable Explanations by Augmenting Static Math Textbooks
• Model Compression in Practice: Lessons Learned from Practitioners Creating On-device Machine Learning Experiences
• KnowledgeVIS: Interpreting Language Models by Comparing Fill-in-the-Blank Prompts
• GAM Coach: Towards Interactive and User-centered Algorithmic Recourse
• Transformer-Driven Inverse Problem Transform for Fast Blind Hyperspectral Image Dehazing
• iScore: Visual Analytics for Interpreting How Language Models Automatically Score Summaries
• MARLens: Understanding Multi-agent Reinforcement Learning for Traffic Signal Control via Visual Analytics
• Spectrogram features for audio and speech analysis
• Explaining Predictions by Characteristic Rules
• Exemplary Natural Images Explain CNN Activations Better than State-of-the-Art Feature Visualization
• WebSHAP: Towards Explaining Any Machine Learning Models Anywhere
• PSEUDo: Interactive Pattern Search in Multivariate Time Series with Locality-Sensitive Hashing and Relevance Feedback
• Why? Why not? When? Visual Explanations of Agent Behavior in Reinforcement Learning
• Unlearning Comparator: A Visual Analytics System for Comparative Evaluation of Machine Unlearning Methods
• Transformer Explainer: Learning LLM Transformers with Interactive Visual Explanation and Experimentation
• Evaluating Interactivity: Toward Automated Assessment of AI-Generated Explorable Explanations
• Towards Accurate RGB-D Saliency Detection with Complementary Attention and Adaptive Integration
• Explainable Adversarial Attacks in Deep Neural Networks Using Activation Profiles