Explaining Deep Face Algorithms through Visualization: A Survey

TL;DR

This survey reviews and analyzes visualization-based explainability algorithms tailored for deep face recognition models, highlighting their nuances, limitations, and practical considerations for AI practitioners.

Contribution

It provides the first comprehensive meta-analysis of face-specific explainability algorithms, assessing their adaptation, effectiveness, and design considerations.

Findings

Identified key challenges in adapting general visualization algorithms to face models.

Revealed insights into the structure and hierarchy of face recognition networks.

Conducted a user study on the utility of various explainability algorithms.

Abstract

Although current deep models for face tasks surpass human performance on some benchmarks, we do not understand how they work. Thus, we cannot predict how it will react to novel inputs, resulting in catastrophic failures and unwanted biases in the algorithms. Explainable AI helps bridge the gap, but currently, there are very few visualization algorithms designed for faces. This work undertakes a first-of-its-kind meta-analysis of explainability algorithms in the face domain. We explore the nuances and caveats of adapting general-purpose visualization algorithms to the face domain, illustrated by computing visualizations on popular face models. We review existing face explainability works and reveal valuable insights into the structure and hierarchy of face networks. We also determine the design considerations for practical face visualizations accessible to AI practitioners by conducting a user study on the utility of various explainability algorithms.