Sexual Violence Among Higher Education Students in the United Kingdom: Results from the Oxford Understanding Relationships, Sex, Power, Abuse and Consent Experiences Study

Sexual violence (SV) experienced by higher education students is a prevalent public health problem. Collecting data on SV through self-report surveys in higher education institutions (HEIs) is essential for estimating the scope of the problem, the first step to adequately resourcing and implementing prevention and response programming and policies. However, in the United Kingdom, data is limited. We used data from the cross-sectional Oxford Understanding Relationships, Sex, Power, Abuse and Consent Experiences survey, administered to all students at a university in the United Kingdom in May 2021 (n = 25,820), to estimate the past year prevalence of SV. We analyzed data from respondents who answered at least one question on SV (n = 1,318) and found that 20.5% of respondents experienced at least one act of attempted or forced sexual touching or rape, and 52.7% of respondents experienced at least one act of sexual harassment (SH). We found that women experienced the highest rates of SV. Attempted forced sexual touching was far more common than forced sexual touching, or rape. Sexist remarks or jokes were the most common act of SH. Most acts of SV took place at the university. These findings reveal that the prevalence of SV in HEIs in the United Kingdom could be far higher than what is experienced in the general population. While this study reflects the context in only one institution, it underlines the need for continued monitoring to develop rigorous, evidence-based, and targeted prevention and response strategies.

Self-supervised learning for using overhead imagery as maps in outdoor range sensor localization

Traditional approaches to outdoor vehicle localization assume a reliable, prior map is available, typically built using the same sensor suite as the on-board sensors used during localization. This work makes a different assumption. It assumes that an overhead image of the workspace is available and utilizes that as a map for use for range-based sensor localization by a vehicle. Here, range-based sensors are radars and lidars. Our motivation is simple, off-the-shelf, publicly available overhead imagery such as Google satellite images can be a ubiquitous, cheap, and powerful tool for vehicle localization when a usable prior sensor map is unavailable, inconvenient, or expensive. The challenge to be addressed is that overhead images are clearly not directly comparable to data from ground range sensors because of their starkly different modalities. We present a learned metric localization method that not only handles the modality difference, but is also cheap to train, learning in a self-supervised fashion without requiring metrically accurate ground truth. By evaluating across multiple real-world datasets, we demonstrate the robustness and versatility of our method for various sensor configurations in cross-modality localization, achieving localization errors on-par with a prior supervised approach while requiring no pixel-wise aligned ground truth for supervision at training. We pay particular attention to the use of millimeter-wave radar, which, owing to its complex interaction with the scene and its immunity to weather and lighting conditions, makes for a compelling and valuable use case.