• Open Access close
• Restricted close
• Open Source close
• Publications close
• Research software close
• Other research products close
• DE close
• English close

Displaying high-quality images on edge devices, such as augmented reality devices, is essential for enhancing the user experience. However, these devices often face power consumption and computing resource limitations, making it challenging to apply many deep learning-based image compression algorithms in this field. Implicit Neural Representation (INR) for image compression is an emerging technology that offers two key benefits compared to cutting-edge autoencoder models: low computational complexity and parameter-free decoding. It also outperforms many traditional and early neural compression methods in terms of quality. In this study, we introduce a new Mixed Autoregressive Model (MARM) to significantly reduce the decoding time for the current INR codec, along with a new synthesis network to enhance reconstruction quality. MARM includes our proposed Autoregressive Upsampler (ARU) blocks, which are highly computationally efficient, and ARM from previous work to balance decoding time and reconstruction quality. We also propose enhancing ARU's performance using a checkerboard two-stage decoding strategy. Moreover, the ratio of different modules can be adjusted to maintain a balance between quality and speed. Comprehensive experiments demonstrate that our method significantly improves computational efficiency while preserving image quality. With different parameter settings, our method can outperform popular AE-based codecs in constrained environments in terms of both quality and decoding time, or achieve state-of-the-art reconstruction quality compared to other INR codecs.

As data gathering technologies are permeating various corners of our lives, a number of stakeholders are attempting to map, track, analyse and define what is happening to our identity, our privacy, or our ways of being social. As notions like privacy, anonymity, data, unlinkability, or pseudonymity are being defined, many of these definitions, while sounding almost the same, shift meaning from discipline to discipline, from context to context, and from one political agenda to the other. In this interview with Marit Hansen, one of the most influential activists for data protection regulation in Germany, and the head of the Independent Centre for Data Protection (ULD) and the Data Protection Commissioner of Schleswig-Holstein, Hansen highlights the way in which her computer science discipline defines its terms and working categories, in a rapidly changing landscape of data gathering technologies. The interview draws heavily from her (co-authored with Andreas Pfitzmann) seminal paper in the computer science field around privacy, anonymity and ‘identity management,’ titled ‘A terminology for talking about privacy by data minimization: Anonymity, unlinkability, undetectability, unobservability, pseudonymity, and identity management’.

Incentivising the implementation of Open Science within the research process remains one of the key barriers towards engagement with open and reproducible research practices. Indeed, many researchers may ask themselves why they should implement Open Science within their research workflow if (from their perspective) such activities are not rewarded. Additionally, there is often a sense that challenges like responsible research evaluations and research assessments with a strong Open Science focus are too large to tackle for a single institution. However, in addition to valuing Open Science from a top-down-level and supporting its researchers with Open Science services, an institution can also fund research projects which implement Open Science themes (e.g., Registered Reports, Citizen Science, Open Data and Code) to incentivise such practices. At our university, researchers are offered the opportunity to apply for annual Open Science Grants of up to 5,000 euros. This lightning talk will cover the Open Science Grants, their motivation, the project requirements with respect to Open Science, our evaluation criteria, the outcomes of the projects themselves and finally our experience as an institution which has implemented this funding pathway. Typically, the evaluation criteria of research funding applications only focus on scientific rigour and expected outcomes. In contrast, the Open Science Grants also require applicants to explain how they will implement Open Science practices and present their overall vision for Open Science in their project. In doing so, these grants ensure that other open aspects of the scholarly process (e.g., pre-registration, sharing data and code openly) are rewarded which would otherwise go unrecognised. After two years providing a moderate amount of money, we see Open Science projects from diverse disciplines (e.g., psychology, social sciences, business studies, linguistics, etc.) at our university. From our experience a key aspect is to open these grants also for all PhD students and early-career researchers from our university, who often lack or cannot apply for any other monetary resources. In this way, we can contribute to a fairer, more inclusive and more equitable distribution of resources within an institution. Moreover, researchers can list the Open Science Grant as an achievement in their CV, an increasingly desired skill demanded in the academic employment market. For institutions, the financial burden is relatively low and thus provides more easily a sustainable monetary solution to ensure researchers are motivated to engage in Open Science. Critically, the Open Science grants chart a new course for reimagining funding opportunities for researchers which encourage, value and reward Open Science practices. This can also stimulate the discussions about evaluation criteria in different research funding programmes or other contexts of research assessments addressed at the Open Science FAIR 2023. We hope that other institutions will follow suit, which could then collectively act as an incubator for an Open Science culture change.

This paper discusses the structural basis and practice of creating a climate policy framework in Nigeria. The roles of existing and planned institutions and legal frameworks are discussed, as well as how they foster or hinder the design of policies and implementation of climate responses at the domestic and international levels. Data has been collected between autumn 2009 and spring 2010. The paper combines literature review with empirical data from interviews of various actors in the Nigerian climate policy process. Various actors including the federal and state governments, civil society and private sector interest groups as well as the Nigerian oil sector are working conjointly on the elaboration of political frameworks and on the realisation of a multitude of individual projects focused on climate change. The paper examines the role of the institutional framework compassing this multitude of stakeholders for the creation of a sustainable climate policy in Nigeria, thereby discussing and referring to literature on institutional governance in the environment sector and institutional change for sustainability. Currently, there is a lack of cohesion between political initiatives and institutions, and a weak implementation of environmental laws and directives. Although becoming a topic of increasing political interest, a policy framework on climate change is still nonexistent for Nigeria – caused not least by diverging lines of interest between participating institutions. The paper shows that once awareness about climate change has been created, there is a danger of the proliferation of institutions and actors addressing the various dimensions of climate change. This proliferation highlights the lack of coordination between various government entities and their struggle to gain/retain influence over the national climate policy, thereby slowing down the design and implementation of responses to mitigate and adapt to climate change. Suggestions are made on how the barriers to climate policy development and implementation in Nigeria can be overcome.

Garcilaso de la Vega el Inca, for several centuries doubtlessly the most discussed and most eminent writer of Andean America in the 16th and 17th centuries, throughout his life set the utmost value on the fact that he descended matrilineally from Atahualpa Yupanqui and from the last Inca emperor, Huayna Capac. Thus, both in his person and in his creative work he combined different cultural worlds in a polylogical way. (1) Two painters boasted that very same Inca descent - they were the last two great masters of the Cuzco school of painting, which over several generations of artists had been an institution of excellent renown and prestige, and whose economic downfall and artistic marginalization was vividly described by the French traveller Paul Mancoy in 1837.(2) While, during the 18th century, Cuzco school paintings were still much cherished and sought after, by the beginning of the following century the elite of Lima regarded them as behind the times and provincial, committed to an 'indigenous' painting style. The artists from up-country - such was the reproach - could not keep up with the modern forms of seeing and creating, as exemplified by European paragons. Yet, just how 'provincial', truly, was this art?

A detailed understanding of the atomistic processes within a lithium ion battery is expected to facilitate the progress of energy storage technology, which is a key element for the exit from nuclear and fossil--fuel energy. Nuclear magnetic resonance (NMR) spectroscopy can provide information about local electronic structure and dynamic properties of battery materials by addressing the nuclear spins of lithium (Li). Support by independent $\textit{ab initio}$ simulations is indispensable to meet the challenges presented by state--of--the--art electrode materials in NMR spectroscopy. In this thesis, the initial steps towards a novel, theoretical multi--scale method for retracing Li NMR experiments on battery materials are presented. First, the theoretical concepts of calculating NMR parameters are summarized. Then, the accuracy of NMR parameter simulations for Li and their dependence on different criteria are benchmarked. On the basis of the high--capacitance, disordered electrode material Li$_4$Ti$_5$O$_{12}$ (LTO), an approach for clustering the NMR quantities according to the local, crystallographic structure is established, reducing the number of $\textit{ab initio}$ NMR calculations. Theoretical sampling of the LTO configuration space demonstrates that the customary one--to--one assignment of experimental observables to crystallographic positions is inaccurate and misleading for complex materials. Finally, a kinetic Monte--Carlo (kMC) model, which simulates the atomistic dynamics, is combined with the NMR autocorrelation function (ACF), which provides the effective observables for spin--alignment echo (SAE) experiments. The simulated, static Li NMR parameters enter the kMC model together with atomic mobility parameters. The kMC sampling of the NMR ACF finally yields observables, which compared to experimental findings confirm the hypothesis of two domains of mobility on separate length and time scales in LTO. Additionally, a theoretical pre--screening approach for organic pH--marker molecules in aqueous solution is outlined. The pre--screening approach is demonstrated for biomedical applications, but might be transferred to energy research applications to study local reaction conditions of transient electrochemical processes. The various results demonstrate the synergy of combining theoretical simulations with NMR experiments. Together theory and experiment amplify the gained knowledge and make additional insights into the basic processes in a battery available. Dissertation, RWTH Aachen University, 2019; Aachen 1 Online-Ressource (xxv, 160 Seiten) : Illustrationen, Diagramme (2019). = Dissertation, RWTH Aachen University, 2019 Published by Aachen