DeepAI AI Chat
Log In Sign Up

Explaining with Counter Visual Attributes and Examples

by   Sadaf Gulshad, et al.
University of Amsterdam

In this paper, we aim to explain the decisions of neural networks by utilizing multimodal information. That is counter-intuitive attributes and counter visual examples which appear when perturbed samples are introduced. Different from previous work on interpreting decisions using saliency maps, text, or visual patches we propose to use attributes and counter-attributes, and examples and counter-examples as part of the visual explanations. When humans explain visual decisions they tend to do so by providing attributes and examples. Hence, inspired by the way of human explanations in this paper we provide attribute-based and example-based explanations. Moreover, humans also tend to explain their visual decisions by adding counter-attributes and counter-examples to explain what is not seen. We introduce directed perturbations in the examples to observe which attribute values change when classifying the examples into the counter classes. This delivers intuitive counter-attributes and counter-examples. Our experiments with both coarse and fine-grained datasets show that attributes provide discriminating and human-understandable intuitive and counter-intuitive explanations.


page 2

page 4

page 6

page 7

page 8


Interactive Label Cleaning with Example-based Explanations

We tackle sequential learning under label noise in applications where a ...

Reflective-Net: Learning from Explanations

Humans possess a remarkable capability to make fast, intuitive decisions...

Grounding Visual Explanations

Existing visual explanation generating agents learn to fluently justify ...

CounterGeDi: A controllable approach to generate polite, detoxified and emotional counterspeech

Recently, many studies have tried to create generation models to assist ...

Towards Inferring Queries from Simple and Partial Provenance Examples

The field of query-by-example aims at inferring queries from output exam...

Quantitative Metrics for Evaluating Explanations of Video DeepFake Detectors

The proliferation of DeepFake technology is a rising challenge in today'...

Squares: A Fast Counter-Based RNG

In this article, we present a new counter-based random number generator ...