The Nonverbal Gap: Toward Affective Computer Vision for Safer and More Equitable Online Dating

A team of researchers has published a provocative vision paper arguing that the computer vision community has a moral responsibility to address safety risks in online dating. The paper, titled 'The Nonverbal Gap: Toward Affective Computer Vision for Safer and More Equitable Online Dating,' contends that current dating platforms strip away the nonverbal cues—such as gaze, facial expression, body posture, and response timing—that humans rely on to signal comfort, disinterest, and consent. This creates a dangerous communication gap with disproportionate consequences for women's safety.

The authors propose a four-part research agenda to bridge this gap using established computer vision (CV) techniques. The agenda focuses on real-time discomfort detection, modeling engagement asymmetry between partners, designing consent-aware interactions, and creating longitudinal interaction summaries. Crucially, they argue this must be done responsibly, requiring purpose-built datasets collected with dyadic consent, fairness evaluations across race and gender, and a commitment to on-device processing to prevent affective data from becoming a surveillance tool. The paper is a direct call to action for the Women in Computer Vision (WICV) community to establish online dating safety as a core research domain before commercial deployment outpaces ethical safeguards.