Contrastive Pretraining for Visual Concept Explanations of Socioeconomic Outcomes

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Obadic, Ivica, Levering, Alex, Pennig, Lars, Oliveira, Dario, Marcos, Diego, Zhu, Xiaoxiang
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916284712615936
author Obadic, Ivica
Levering, Alex
Pennig, Lars
Oliveira, Dario
Marcos, Diego
Zhu, Xiaoxiang
author_facet Obadic, Ivica
Levering, Alex
Pennig, Lars
Oliveira, Dario
Marcos, Diego
Zhu, Xiaoxiang
contents Predicting socioeconomic indicators from satellite imagery with deep learning has become an increasingly popular research direction. Post-hoc concept-based explanations can be an important step towards broader adoption of these models in policy-making as they enable the interpretation of socioeconomic outcomes based on visual concepts that are intuitive to humans. In this paper, we study the interplay between representation learning using an additional task-specific contrastive loss and post-hoc concept explainability for socioeconomic studies. Our results on two different geographical locations and tasks indicate that the task-specific pretraining imposes a continuous ordering of the latent space embeddings according to the socioeconomic outcomes. This improves the model's interpretability as it enables the latent space of the model to associate concepts encoding typical urban and natural area patterns with continuous intervals of socioeconomic outcomes. Further, we illustrate how analyzing the model's conceptual sensitivity for the intervals of socioeconomic outcomes can shed light on new insights for urban studies.
format Preprint
id arxiv_https___arxiv_org_abs_2404_09768
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Contrastive Pretraining for Visual Concept Explanations of Socioeconomic Outcomes
Obadic, Ivica
Levering, Alex
Pennig, Lars
Oliveira, Dario
Marcos, Diego
Zhu, Xiaoxiang
Computer Vision and Pattern Recognition
Predicting socioeconomic indicators from satellite imagery with deep learning has become an increasingly popular research direction. Post-hoc concept-based explanations can be an important step towards broader adoption of these models in policy-making as they enable the interpretation of socioeconomic outcomes based on visual concepts that are intuitive to humans. In this paper, we study the interplay between representation learning using an additional task-specific contrastive loss and post-hoc concept explainability for socioeconomic studies. Our results on two different geographical locations and tasks indicate that the task-specific pretraining imposes a continuous ordering of the latent space embeddings according to the socioeconomic outcomes. This improves the model's interpretability as it enables the latent space of the model to associate concepts encoding typical urban and natural area patterns with continuous intervals of socioeconomic outcomes. Further, we illustrate how analyzing the model's conceptual sensitivity for the intervals of socioeconomic outcomes can shed light on new insights for urban studies.
title Contrastive Pretraining for Visual Concept Explanations of Socioeconomic Outcomes
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2404.09768