Skeleton sequences are lightweight and compact and thus are ideal candidates for action recognition on edge devices. Recent skeleton-based action recognition methods extract features from 3-D joint coordinates as spatial–temporal cues, using these representations in a graph neural network for feature fusion to boost recognition performance. The use of first-and second-order features, that is, joint and bone representations, has led to high accuracy. Nonetheless, many models are still confused by actions that have similar motion trajectories. To address these issues, we propose fusing higher-order features in the form of angular encoding (AGE) into modern architectures to robustly capture the relationships between joints and body parts. This simple fusion with popular spatial–temporal graph neural networks achieves new state-of-the-art accuracy in two large benchmarks, including NTU60 and NTU120, while employing fewer parameters and reduced run time. Our source code is publicly available at: https://github.com/ZhenyueQin/Angular-Skeleton-Encoding.

The detection and localization of highly realistic deepfake audio-visual content are challenging even for the most advanced state-of-the-art methods. While most of the research efforts in this domain are focused on detecting high-quality deepfake images and videos, only a few works address the problem of the localization of small segments of audio-visual manipulations embedded in real videos. In this research, we emulate the process of such content generation and propose the AV-Deepfake1M dataset. The dataset contains content-driven (i) video manipulations, (ii) audio manipulations, and (iii) audio-visual manipulations for more than 2K subjects resulting in a total of more than 1M videos. The paper provides a thorough description of the proposed data generation pipeline accompanied by a rigorous analysis of the quality of the generated data. The comprehensive benchmark of the proposed dataset utilizing state-of-the-art deepfake detection and localization methods indicates a significant drop in performance compared to previous datasets. The proposed dataset will play a vital role in building the next-generation deepfake localization methods. The dataset and associated code are available at https://github.com/ControlNet/AV-Deepfake1M.

Given the prevalence of depression worldwide and its major impact on society, several studies employed artificial intelligence modelling to automatically detect and assess depression. However, interpretation of these models and cues are rarely discussed in detail in the AI community, but have received increased attention lately. In this article, we aim to analyse the commonly selected features using a proposed framework of several feature selection methods and their effect on the classification results, which will provide an interpretation of the depression detection model. The developed framework aggregates and selects the most promising features for modelling depression detection from 38 feature selection algorithms of different categories. Using three real-world depression datasets, 902 behavioural cues were extracted from speech behaviour, speech prosody, eye movement and head pose. To verify the generalisability of the proposed framework, we applied the entire process to depression datasets individually and when combined. The results from the proposed framework showed that speech behaviour features (e.g. pauses) are the most distinctive features of the depression detection model. From the speech prosody modality, the strongest feature groups were F0, HNR, formants, and MFCC, while for the eye activity modality they were left-right eye movement and gaze direction, and for the head modality it was yaw head movement. Modelling depression detection using the selected features (even though there are only 9 features) outperformed using all features in all the individual and combined datasets. Our feature selection framework did not only provide an interpretation of the model, but was also able to produce a higher accuracy of depression detection with a small number of features in varied datasets. This could help to reduce the processing time needed to extract features and creating the model.

Research background: Generative artificial intelligence (AI) and machine learning algorithms support industrial Internet of Things (IoT)-based big data and enterprise asset management in multiphysics simulation environments by industrial big data processing, modeling, and monitoring, enabling business organizational and managerial practices. Machine learning-based decision support and edge generative AI sensing systems can reduce persistent labor shortages and job vacancies and power productivity growth and labor market dynamics, shaping career pathways and facilitating occupational transitions by skill gap identification and labor-intensive manufacturing job automation by path planning and spatial cognition algorithms, furthering theoretical implications for management sciences. Generative AI fintech, machine learning algorithms, and behavioral analytics can assist multi-layered payment and transaction processing screening with regard to authorized push payment, account takeover, and synthetic identity frauds, flagging suspicious activities and combating economic crimes by rigorous verification processes. Purpose of the article: We show that edge device management functionalities of cloud industrial IoT and virtual robotic simulation technologies configure plant production and route planning processes across cyber-physical production and industrial automation systems in multi-cloud immersive 3D environments, leading to tangible business outcomes by reinforcement learning and convolutional neural networks. Labor-augmenting automation and generative AI technologies can impact employment participation, increase wage and wealth inequality, and lead to potential job displacement and massive labor market disruptions. The deep learning capabilities of generative AI fintech in terms of adaptive behavioral analytics and credit scoring mechanisms can enhance financial transaction behaviors and algorithmic trading returns, identify fraudulent payment transactions swiftly, and improve financial forecasts, leading to customized investment recommendations and well-informed financial decisions. Methods: Machine learning-based study selection process and text mining systematic review management software and tools leveraged include Abstrackr, CADIMA, Colandr, DistillerSR, EPPI-Reviewer, JBI SUMARI, METAGEAR package for R, SluRp, and SWIFT-Active Screener. Such reference management systems are harnessed for methodologically rigorous evidence synthesis, study selection and characteristic extraction, predictive document classification, machine learning-based citation and record screening, bias assessment, article retrieval automation, and document classification and prioritization. Findings & value added: Industrial IoT and 3D augmented reality technologies can create business value by streamlining virtual product and remote asset management across extended reality-based navigation and robotic autonomous systems in smart factory environments by generative AI and machine learning algorithms, articulating business organizational level and theory of management implications. 3D simulation and operational modeling tools can execute and complete complex cognitive task-oriented and knowledge economy jobs, producing first-rate quality outputs swiftly while leading to unemployment spells, labor market disruptions, job displacement losses, and reduced earnings by machine learning clustering and spatial cognition algorithms. Generative AI decentralized finance, interoperable blockchain networks, cash flow management tools, and asset tokenization can mitigate fraud risks, enable digital fund and crypto investing servicing, and automate treasury operations by integrating real-time payment capabilities, routing and configurable workflows, and lending and payment technologies.