Situated Cameras, Situated Knowledges: Towards an Egocentric Epistemology for Computer Vision

TL;DR

This paper proposes an 'Egocentric Epistemology' framework for computer vision, integrating feminist epistemology to emphasize human perspectives and qualitative methods alongside traditional performance metrics.

Contribution

It introduces a novel perspective linking feminist epistemology with egocentric computer vision, advocating for human-centric qualitative approaches.

Findings

Highlights the importance of human perspectives in CV

Suggests qualitative methods complement performance benchmarks

Connects feminist epistemology with technical CV approaches

Abstract

In her influential 1988 paper, Situated Knowledges, Donna Haraway uses vision and perspective as a metaphor to discuss scientific knowledge. Today, egocentric computer vision discusses many of the same issues, except in a literal vision context. In this short position paper, we collapse that metaphor, and explore the interactions between feminist epistemology and egocentric CV as "Egocentric Epistemology." Using this framework, we argue for the use of qualitative, human-centric methods as a complement to performance benchmarks, to center both the literal and metaphorical perspective of human crowd workers in CV.