Partial Success in Closing the Gap between Human and Machine Vision | Robust Machine Learning – Max Planck Institute for Intelligent Systems

Institute Homepage

Institute Homepage Sign In

Research Overview

Comparative Vision Science

Does CLIP's Generalization Performance Mainly Stem from High Train-Test Similarity?

In Search of Forgotten Domain Generalization

Partial Success in Closing the Gap between Human and Machine Vision

Foundations of Machine Learning

Cross-Entropy Is All You Need To Invert the Data Generating Process

Contrastive Learning Inverts the Data Generating Process

Interaction Asymmetry: A General Principle for Learning Composable Abstractions

Mechanistic Interpretability

Measuring Per-Unit Interpretability at Scale Without Humans

Exemplary Natural Images explain CNN Activations better than State-Of-The-Art Feature Visualizations

How Well do Feature Visualizations Support Causal Understanding of CNN Activations?

Robust Machine Learning

Partial Success in Closing the Gap between Human and Machine Vision

Error consistency — The difference in error consistency between human-human and human-machine remains high despite the progress that foundation models have made in closing the gap.

We tested human and machine vision across diverse OOD datasets. While machines are closing the robustness gap with humans and improving with larger training datasets, significant behavioral differences and error consistency gaps remain, highlighting both progress and areas for further improvement~[

].