While the significant advancements have made in the generation of deepfakes using deep learning technologies, its misuse is a well-known issue now. Deepfakes can cause severe security and privacy issues as they can be used to impersonate a person's identity in a video by replacing his/her face with another person's face. Recently, a new problem of generating synthesized human voice of a person is emerging, where AI-based deep learning models can synthesize any person's voice requiring just a few seconds of audio. With the emerging threat of impersonation attacks using deepfake audios and videos, a new generation of deepfake detectors is needed to focus on both video and audio collectively. A large amount of good quality dataset is typically required to capture the real-world scenarios to develop a competent deepfake detector. Existing deepfake datasets either contain deepfake videos or audios, which are racially biased as well. Hence, there is a crucial need for creating a good video as well as audio deepfake dataset, which can be used to detect audio and video deepfake simultaneously. To fill this gap, we propose a novel Audio-Video Deepfake dataset (FakeAVCeleb) that contains not only deepfake videos but also respective synthesized lip-synced fake audios. We generate this dataset using the current most popular deepfake generation methods. We selected real YouTube videos of celebrities with four racial backgrounds (Caucasian, Black, East Asian, and South Asian) to develop a more realistic multimodal dataset that addresses racial bias, and further help develop multimodal deepfake detectors. We performed several experiments using state-of-the-art detection methods to evaluate our deepfake dataset and demonstrate the challenges and usefulness of our multimodal Audio-Video deepfake dataset.Preprint. Under review.
Three highway bridges in the Canadian province of Manitoba are being monitored continuously not only for their long-term performance but also for bridge weighing-in-motion (BWIM). Data collected for the BWIM study has led to some observations that have far-reaching consequences about the design and evaluation loads for highway bridges. This paper presents the well-known concept of equivalent base length, Bm, as a useful tool for comparing trucks with different axle weight and spacing configurations as they influence load effects in all bridges. It is discussed that the statistics of gross vehicle weights (GVWs), W, collected over a one-month period is not significantly different from that for the GVW data collected over a longer period. A rational method concludes that the value of W for the CL-W Truck, the design live load specified by the Canadian Highway Bridge Design Code, is 555 kN for Manitoba. The observed truck data in Manitoba presented on the W–Bm space is found to be similar to that collected in the Canadian province of Ontario more than four decades ago. It was also found that the multi-presence factors, accounting for the presence of side-by-side trucks in two-lane bridges, specified in North American bridge design and evaluation codes are somewhat conservative.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.