Jean-Luc Picard at Touch\'e 2023: Comparing Image Generation, Stance Detection and Feature Matching for Image Retrieval for Arguments

TL;DR

This paper evaluates various image retrieval pipelines combining image generation, stance detection, preselection, and feature matching for argument-based image retrieval, comparing their performance to a baseline.

Contribution

It introduces multiple pipeline configurations for argument image retrieval and compares their effectiveness against a baseline.

Findings

All pipelines performed similarly to the baseline.

Different pipeline layouts showed comparable retrieval performance.

The study highlights the challenges in improving argument image retrieval accuracy.

Abstract

Participating in the shared task "Image Retrieval for arguments", we used different pipelines for image retrieval containing Image Generation, Stance Detection, Preselection and Feature Matching. We submitted four different runs with different pipeline layout and compare them to given baseline. Our pipelines perform similarly to the baseline.