Log In Sign Up

Enabling Verification of Deep Neural Networks in Perception Tasks Using Fuzzy Logic and Concept Embeddings

by   Gesina Schwalbe, et al.

One major drawback of deep convolutional neural networks (CNNs) for use in safety critical applications is their black-box nature. This makes it hard to verify or monitor complex, symbolic requirements on already trained computer vision CNNs. In this work, we present a simple, yet effective, approach to verify that a CNN complies with symbolic predicate logic rules which relate visual concepts. It is the first that (1) does not modify the CNN, (2) may use visual concepts that are no CNN in- or output feature, and (3) can leverage continuous CNN confidence outputs. To achieve this, we newly combine methods from explainable artificial intelligence and logic: First, using supervised concept embedding analysis, the output of a CNN is post-hoc enriched by concept outputs. Second, rules from prior knowledge are modelled as truth functions that accept the CNN outputs, and can be evaluated with little computational overhead. We here investigate the use of fuzzy logic, i.e., continuous truth values, and of proper output calibration, which both theoretically and practically show slight benefits. Applicability is demonstrated on state-of-the-art object detectors for three verification use-cases, where monitoring of rule breaches can reveal detection errors.


page 7

page 17


Entropy-based Logic Explanations of Neural Networks

Explainable artificial intelligence has rapidly emerged since lawmakers ...

Visualizing Convolutional Neural Networks to Improve Decision Support for Skin Lesion Classification

Because of their state-of-the-art performance in computer vision, CNNs a...

From Black-box to White-box: Examining Confidence Calibration under different Conditions

Confidence calibration is a major concern when applying artificial neura...

Towards a Measure of Trustworthiness to Evaluate CNNs During Operation

Due to black box nature of Convolutional neural networks (CNNs), the con...

Harnessing Deep Neural Networks with Logic Rules

Combining deep neural networks with structured logic rules is desirable ...

Verification of Size Invariance in DNN Activations using Concept Embeddings

The benefits of deep neural networks (DNNs) have become of interest for ...

Cognitive Explainers of Graph Neural Networks Based on Medical Concepts

Although deep neural networks (DNN) have achieved state-of-the-art perfo...