Falls are a major cause of injuries and deaths among older adults worldwide. Accurate fall detection can help reduce potential injuries and additional health complications. Different types of video modalities can be used in a home setting to detect falls, including RGB, Infrared, and Thermal cameras. Anomaly detection frameworks using autoencoders and their variants can be used for fall detection due to the data imbalance that arises from the rarity and diversity of falls. However, the use of reconstruction error in autoencoders can limit the application of networks' structures that propagate information. In this paper, we propose a new multi-objective loss function called Temporal Shift, which aims to predict both future and reconstructed frames within a window of sequential frames. The proposed loss function is evaluated on a semi-naturalistic fall detection dataset containing multiple camera modalities. The autoencoders were trained on normal activities of daily living (ADL) performed by older adults and tested on ADLs and falls performed by young adults. Temporal shift shows significant improvement to a baseline 3D Convolutional autoencoder, an attention U-Net CAE, and a multi-modal neural network. The greatest improvement was observed in an attention U-Net model improving by 0.20 AUC ROC for a single camera when compared to reconstruction alone. With significant improvement across different models, this approach has the potential to be widely adopted and improve anomaly detection capabilities in other settings besides fall detection.

Fast disaster impact reporting is crucial in planning humanitarian assistance. Large Language Models (LLMs) are well known for their ability to write coherent text and fulfill a variety of tasks relevant to impact reporting, such as question answering or text summarization. However, LLMs are constrained by the knowledge within their training data and are prone to generating inaccurate, or "hallucinated", information. To address this, we introduce a sophisticated pipeline embodied in our tool FloodBrain (floodbrain.com), specialized in generating flood disaster impact reports by extracting and curating information from the web. Our pipeline assimilates information from web search results to produce detailed and accurate reports on flood events. We test different LLMs as backbones in our tool and compare their generated reports to human-written reports on different metrics. Similar to other studies, we find a notable correlation between the scores assigned by GPT-4 and the scores given by human evaluators when comparing our generated reports to human-authored ones. Additionally, we conduct an ablation study to test our single pipeline components and their relevancy for the final reports. With our tool, we aim to advance the use of LLMs for disaster impact reporting and reduce the time for coordination of humanitarian efforts in the wake of flood disasters.

We outline an unsupervised method for temporal rank ordering of sets of historical documents, namely American State of the Union Addresses and DEEDS, a corpus of medieval English property transfer documents. Our method relies upon effectively capturing the gradual change in word usage via a bandwidth estimate for the non-parametric Generalized Linear Models (Fan, Heckman, and Wand, 1995). The number of possible rank orders needed to search through possible cost functions related to the bandwidth can be quite large, even for a small set of documents. We tackle this problem of combinatorial optimization using the Simulated Annealing algorithm, which allows us to obtain the optimal document temporal orders. Our rank ordering method significantly improved the temporal sequencing of both corpora compared to a randomly sequenced baseline. This unsupervised approach should enable the temporal ordering of undated document sets.

Neural network models are vulnerable to adversarial examples, and adversarial transferability further increases the risk of adversarial attacks. Current methods based on transferability often rely on substitute models, which can be impractical and costly in real-world scenarios due to the unavailability of training data and the victim model's structural details. In this paper, we propose a novel approach that directly constructs adversarial examples by extracting transferable features across various tasks. Our key insight is that adversarial transferability can extend across different tasks. Specifically, we train a sequence-to-sequence generative model named CT-GAT using adversarial sample data collected from multiple tasks to acquire universal adversarial features and generate adversarial examples for different tasks. We conduct experiments on ten distinct datasets, and the results demonstrate that our method achieves superior attack performance with small cost.

While large language models (LLMs) can answer many questions correctly, they can also hallucinate and give wrong answers. Wikidata, with its over 12 billion facts, can be used to ground LLMs to improve their factuality. This paper presents WikiWebQuestions, a high-quality question answering benchmark for Wikidata. Ported over from WebQuestions for Freebase, it consists of real-world data with SPARQL annotation. This paper presents a few-shot sequence-to-sequence semantic parser for Wikidata. We modify SPARQL to use the unique domain and property names instead of their IDs. We train the parser to use either the results from an entity linker or mentions in the query. We fine-tune LLaMA by adding the few-shot training data to that used to fine-tune Alpaca. Our experimental results demonstrate the effectiveness of this methodology, establishing a strong baseline of 76% and 65% answer accuracy in the dev and test sets of WikiWebQuestions, respectively. By pairing our semantic parser with GPT-3, we combine verifiable results with qualified GPT-3 guesses to provide useful answers to 96% of the questions in dev. We also show that our method outperforms the state-of-the-art for the QALD-7 Wikidata dataset by 3.6% in F1 score.

A primary goal in strategic classification is to learn decision rules which are robust to strategic input manipulation. Earlier works assume that strategic responses are known; while some recent works address the important challenge of unknown responses, they exclusively study sequential settings which allow multiple model deployments over time. But there are many domains$\unicode{x2014}$particularly in public policy, a common motivating use-case$\unicode{x2014}$where multiple deployments are unrealistic, or where even a single bad round is undesirable. To address this gap, we initiate the study of strategic classification under unknown responses in the one-shot setting, which requires committing to a single classifier once. Focusing on the users' cost function as the source of uncertainty, we begin by proving that for a broad class of costs, even a small mis-estimation of the true cost can entail arbitrarily low accuracy in the worst case. In light of this, we frame the one-shot task as a minimax problem, with the goal of identifying the classifier with the smallest worst-case risk over an uncertainty set of possible costs. Our main contribution is efficient algorithms for both the full-batch and stochastic settings, which we prove converge (offline) to the minimax optimal solution at the dimension-independent rate of $\tilde{\mathcal{O}}(T^{-\frac{1}{2}})$. Our analysis reveals important structure stemming from the strategic nature of user responses, particularly the importance of dual norm regularization with respect to the cost function.

Earlier last week, just around the time when the driver of Rishi Sunak's armoured Jaguar might have been thinking about typing "Bletchley Park" into the limousine's satnav, Joe Biden was in the White House putting his signature on a new executive order "on the safe, secure, and trustworthy development and use of artificial intelligence". In a mere 20,000 words, or thereabouts, the order directs an innumerable number of federal agencies and government departments that oversee "everything from housing to health to national security to create standards and regulations for the use or oversight of AI". These bodies are required to develop guidance on the responsible use of AI in areas such as criminal justice, education, healthcare, housing and labour, "with a focus on protecting Americans' civil rights and liberties". Within No 10, though, there might have been some infuriated spin doctors. After all, the main purpose of the Bletchley Park AI safety summit was to hype the prime minister's claim to "global leadership" in this matter, and here was bloody Biden announcing tangible plans actually to do something about the technology rather than just fostering lofty "declarations".

After three weeks of a punishing Israeli bombardment of Gaza, Israel is still refusing to allow international journalists in. News outlets and audiences are entirely reliant on local Palestinian reporters, who risk their lives to provide a window into the war. Meenakshi Ravi reports on how Israelis are documenting and sharing the evidence online. Tariq Nafi examines Israel's use of AI-powered surveillance in Hebron, which has entrenched the Israeli government's control over Palestinians.

Rep. Nancy Mace, R-S.C., says Congress should not be combining Israel and Ukraine wars together in aid package on'Your World.' Rep. Nancy Mace, R-S.C., is calling for solutions to the wide array of dangers posed by online content falsified using Artificial Intelligence (AI) – known as "deepfakes." "These things are only going to become more prevalent if we don't start discussing the problem and talking to AI experts on how to address deepfakes now and in the future," Mace told Fox News Digital in a Friday interview. She hopes to get those answers in next week's hearing on AI deepfakes by the House Oversight's Subcommittee on Cybersecurity, Information Technology, and Government Innovation – which Mace chairs. Mace said she hopes the expert witnesses at the Wednesday hearing will "share some of the more egregious examples" of AI deepfakes being used, like the prevalence of obscene AI generated images and video. Rep. Nancy Mace, R-S.C., is chairing a House Oversight subcommittee hearing on AI (Tom Williams/CQ-Roll Call, Inc via Getty Images) "Ninety percent of AI deepfakes are pornographic in nature," Mace said, listing off the dangers of AI-faked content.

Predictive models used for AI-based decision support are generally not designed for transparency. Although they operate in critical situations such as, e.g., medicine or defence, they are limited to only presenting a probable outcome (David Gunning, 2017; Ribeiro et al., 2016), which can lead to either misuse (based on user reliance being higher than appropriate) or disuse (due to users having less reliance than appropriate) (Alvarado-Valencia & Barrero, 2014; Buçinca et al., 2020). Due to the lack of transparency, predictions from this type of model often require an explanation. In explainable artificial intelligence (XAI), the goal is to create methods that help human users identify when to trust a prediction and when not to, such as an erroneous prediction in a medical diagnosis (Marx et al., 2023). An explanation should reveal the strengths and weaknesses of the underlying model to communicate how they will behave in the future (David Gunning, 2017; Dimanov et al., 2020). There are two main categories of explanations: local explanations, which present information about the reasons for individual predictions, and global explanations, which provide information about the general behaviour of the model (Guidotti et al., 2018b; Moradi & Samwald, 2021; Martens & Foster, 2014).