26199 research outputs found
Sort by
Body-part Tubelet Transformer for Human-Related Crime Classification
Detecting human-related crimes from surveillance videos poses an increasingly difficult challenge, especially when confronted with human actions that are relatively similar. In this work, we propose a transformer-based model that induces bias through the incorporation of a Tubelet embedder module-a 3D convolutional layer. The aim is to capture spatiotemporal embeddings from skeletal trajectories extracted from videos using 3D convolutional operations. Our experiments are conducted on the Human-Related Crime dataset, revealing that the use of tubelet embeddings maintains competitive performance (49% accuracy) to the state-of-the-art, while considerably reducing the computational complexity of the model
Beyond Birth: Exploring the Complexities & Potential Misuse of Artificial Womb Technology
Artificial Womb Technology (AWT) promises revolutionary advancements in infertility treatments and neonatal care for premature infants. This paper conducts an examination of AWT's ethical, legal, social, environmental, economic, and cybersecurity implications. Methodologically, it integrates an extensive review of existing literature with a theoretical analysis of potential misuse patterns. The ethical discussion addresses concern about the commodification of human life, parental rights, and the disruption of natural birth processes. Legal challenges focus on the necessity for new regulations to govern usage, consent, and responsibility. Social implications highlight the potential impact on family dynamics and the risk of worsening existing inequalities, while the environmental considerations are centered on the resource demands, overpopulation and waste associated with AWT. Economically, AWT introduces new market opportunities while also raising concerns about commercialization and profit-driven misuse. Cybersecurity emerges as a critical, overlooked issue due to the sensitive nature of the data involved and the severe consequences of potential breaches. The theoretical analysis highlights the historical misuse patterns in digital health technologies, reinforcing the urgent need for stringent guidelines and policies to ensure the responsible implementation of AWT
3D millimeter-Wave Sensing vs Ultra-Wideband Positioning
Indoor positioning and sensing using millimeter-wave (mmWave) and Ultra-Wideband (UWB) technologies have garnered significant attention in the literature. While extensive research exists on 2 D positioning with these technologies, a notable gap remains in addressing 3D positioning. Existing studies predominantly focus on the horizontal plane of localization, overlooking the necessity of vertical dimension integration. This paper identifies this and directly compares mmWave and UWB for 3D localisation. Addressing this gap, our work conducts a comparative analysis of 3D sensing with mmWave and 3D positioning with UWB technologies, evaluating the accuracy, robustness, efficiency, and associated challenges. This research underscores the need for future investigations to explore and assess the performance of these technologies in three dimensions
Of bits and pieces of EU law in territories. The many shades of European integration.
The very first provision of the Treaty on the European Union (TEU) clearly states that the Contracting Parties to the Treaty establish among themselves a Union, ‘on which the Member States confer competences to attain objectives they have in common’. To the extent that the attainment of ‘objectives Member States have in common’ may have been diverse or differentiated throughout the evolution of the European project, as developed in the present edited volume, this chapter intends to provide a renewed analysis of key manifestations of differentiated integration in today’s European Union (EU) by reference to Member State territories (the sum of their various parts or areas falling under their jurisdiction) and their possible impact on the re(definition) of EU membership. It is argued that notwithstanding initial delimitations in the EU Treaties, the concept of EU membership and the principle of adherence to the EU acquis have taken on renewed dimensions during the last decades of the ‘ever-closer Union’. Throughout the existence of the EU, Member States may have been allowed to expressly opt out from specific areas or policies in initial phases of the European project’s construction, while in later stages, newer Member States may be ‘exempted’ from EU provisions and/or principles at the core of EU law, as a matter of ‘necessity’, benefiting from temporary transitional regimes and/or more permanent derogations/exemptions/limitations. At the same time, specific territories, internal or external to the Union, exceptionally remain de facto and/or de jure integrated in various EU policies. These multiple shades of European integration are probably most perceivable among some of the territories of Member States and parts of the United Kingdom (UK) territory particularly, and through their complex relations with EU law. At the same time, the withdrawal clause from the Union has now been acted upon in the context of Brexit, redesigning the landscape for the UK territory, its relationship with the EU and with Member States in a set of complex neighbouring relations
The Quaker Moons of Lancashire
In April 1658, a Lancastrian currier named John Moon claimed to receive a revelation from Jesus Christ that affirmed many tenets taught by the English Dissenter George Fox. Moon, along with his parents, brother and uncle, joined the Religious Society of Friends in 1653, a year after Fox’s vision on Pendle Hill, not far from the Moon’s homeland on the plains of the Fylde. This portion of the Moon family received instant reprisal for converting to the new movement. However, imprisonments, fines and the seizure of his possessions did not stop John from publishing several works affirming his newfound beliefs. Still, scholars remain in dispute over parts of the history of the Moon family’s conversion. This article investigates the early history of Quakerism in Lancashire with special attention given to John Moon’s literary works which may warrant his recognition as a Quaker leader key to the propagation of the movement
Constraint-induced movement therapy following stroke: a commentary
Over 113000 patients present with stroke each year in the UK, the societal cost of which is approximately £26 billion annually, with £20.6 billion attributed to care after stroke. Approximately 70% of stroke survivors suffer from impaired arm function, with recovery patterns heavily influenced by initial motor weakness. Constraint-induced movement therapy has demonstrated some potential in rehabilitating this dysfunction, when finger extension is preserved. Therefore, constraint-induced movement therapy is now recommended in national healthcare guidelines. Systematic reviews of constraint-induced movement therapy have varied in their sample groups, focusing on acute, subacute and chronic strokes, with varying delivery protocols. A systematic review was undertaken in 2023 with the aim of identifying the efficiency of constraint induced movement therapy in patients with preserved finger extension, as well as the optimum protocol for delivery. This commentary critically appraises the methods used in the review and expands on the findings in the context of clinical practice
New insights on supernova remnants and HII regions in M82
The nearby (d = 3.6 Mpc) starburst galaxy M82 has been studied for several decades by very long baseline interferometry (VLBI) networks such as e-MERLIN and the European VLBI Network (EVN). The numerous supernova remnants (SNRs), H ii regions and other exotic transients make it a perfect laboratory for studying stellar evolution and the interstellar medium (ISM). Its proximity provides a linear resolution of 17 pc/arcsec, enabling decadal-time-scale variability and morphology studies of the tens of compact radio sources. In these proceedings, we describe the use of new techniques developed in the last ten years that provide deeper, more robust imaging, enable in-band spectral index mapping, and allow wider fields of view to be imaged to find new radio sources
Exploring rehabilitation technology adoption in clinical practice in the UK: introducing the RiTe model.
Background: Digital health technologies (DHT) comprise a broad range of applications such as telehealth, wearable devices and smart-phone and tablet applications (apps). However, whilst national and international policies present ambitious plans for DHT to revolutionize healthcare, there has been little consideration of how they can be successfully integrated into healthcare systems and processes. This is important as many reports show that even well designed DHT fail to be adopted or are quickly abandoned in clinical practice, meaning that their potential to transform healthcare and patient outcomes are lost. Explicit recognition of the key challenges and facilitators of rehabilitation technology implementation in clinical practice is vital to ensure that technology adoption can be planned and evaluated and that technologies are selected, utilized and adopted sustainably. To our knowledge, there are no rehabilitation technology specific implementation models that are available to support this process. Therefore, the aim of this work was to develop a comprehensive model to guide rehabilitation technology implementation. To do this, it is important to understand both the theories, frameworks and models that are used to
support rehabilitation technology adoption and to capture the lived experience of people involved in rehabilitation technology in clinical rehabilitation
Methods: A multi-faceted methodology was utilized. Systematic, scoping and hermeneutic reviews were undertaken to identify existing implementation models, theories, frameworks and individual factors that have been used to underpin adoption of rehabilitation technology in neurological conditions and stroke. Stakeholder analysis identified a range of people and roles who had influence upon technology adoption. These comprised clinicians, patients, managers, organization leads, innovation groups, digital developers who produced the technology, Information technology and information governance roles. After ethical approval, individuals based in the UK were interviewed using a schedule developed from Normalization Process Theory. Interviews were transcribed and analyzed using constant comparison techniques. Codes and themes were generated and group into overarching areas, then triangulated with findings from the literature reviews. These were then assimilated into a final model.
Results: The literature reviews identified over 10 broad models and a range of individual reviews and studies of factors influencing rehabilitation technology implementation. Qualitative interviews generated more than 300 codes, which were assimilated into initial 15 themes and then grouped into 5 overarching themes, supported by existing models and factors found in the literature. These themes comprised factors centered around: the Team, Organization, Users, Technology and Evidence. This produced the Rehabilitation Technologies (RiTe) Implementation Model. Associated guidance and toolkits to support the use of the model in rehabilitation technology are currently being produced and will be housed on our website with the model: advancingrehab.com
Conclusions: To our knowledge, the RiTe model is the first specific implementation model to support rehabilitation technology adoption into clinical practice. We hope that it will be used to support and evaluate rehabilitation technology use, avoid costly failed adoption and enable more patients to benefit from technologies in their rehabilitation.
Whilst this model was generated solely from UK focused interviews, we anticipate many of the factors will have relevance in other healthcare settings, most notably those with a universal healthcare structure.
We are keen to work with global partners to test and refine the RiTe model for use in other healthcare settings and will continue to refine the model based upon user feedback