Published in IGARSS, 2023
TL;DR We surveyed AI experts and found varying attitudes across gender and age for the main principles of Responsible AI, such as transparency and sustainability.
Download here
Published in Nature Humanities and Social Sciences Communications 10 (1), 1-9, 2023
TL;DR While many modern corporations proclaim their commitment to Corporate Social Responsibility (CSR), allegations of “greenwashing” are not uncommon. This study employs Glassdoor reviews to quantify an underexplored facet of CSR – internal sustainability. This broad term encapsulates various policies targeted towards internal stakeholders, such as advocating for a balanced work-life ecosystem, prioritizing gender equality and diversity, and fostering a harassment-free work environment.
Download here
Published in Nature Scientific Reports 13 (1), 1603, 2023
TL;DR Our study unveils the under-recognized “positive stress” phenomenon within employee reviews across S&P 500 companies. Notably, we highlight its positive association with long-term company stock growth.
Download here
Published in Nature Scientific Reports 13 (1), 6711, 2023
TL;DR Our study challenges the traditional understanding of online news circulation, demonstrating that it is not just individual characteristics and social media algorithms at play, but also the influence of localized offline self-segregation (the Big Sort) in the US. The study underscores that the diffusion of news, especially misinformation, is significantly influenced by state-level personality, cultural factors, and education levels, highlighting the geographical nature of online interactions.
Download here
Published in 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2022
TL;DR We present our pioneering findings that human stress levels can be determined with a certain degree of precision through abdominal sound analysis.
Download here
Published in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021
TL;DR This paper was our assessment of state-of-the-art semantic segmentation deep learning models for the crucial task land cover mapping.
Download here
Published in The Royal Society Open Science, 2021
TL;DR We show that COVID dreams feature both common and specific symptom mentions, when compared to waking life COVID experiences. E.g., expressions in waking life described realistic symptoms (e.g. nasal pain, SARS, H1N1); while those in dreaming life reflected either conditions unrelated to the virus (e.g. maggots, deformities, snake bites), or conditions of surreal nature (e.g. teeth falling out, body crumbling into sand).
Download here
Published in Proceedings of the ACM on Human-Computer Interaction, 2021
TL;DR Our study demonstrates the utility of open Sentinel-2 satellite imagery in assessing the vitality of diverse city neighborhoods. This reinforces renowned urban theory by Jane Jacobs, identifying key elements that foster pedestrian activity and vibrant neighborhoods.
Download here