organization

University of Sheffield

Country: United Kingdom
Funder (3)
Top 100 values are shown in the filters
Results number
arrow_drop_down
3,677 Projects, page 1 of 736
  • Funder: UKRI Project Code: 2588088
    Partners: University of Sheffield

    Speech and natural language data will often contain many long-range dependencies, where there is some statistical correlation between words or speech that are spatially distant [1]. For example, in conversational data, the interactions between multiple speakers can produce sequences of utterances that are logically connected and reference information from far in the past. As a consequence, systems that have a greater capacity to model and learn these relationships will show greater predictive performance. Despite this, the majority of the current Automatic Speech Recognition (ASR) systems are designed to model speech as an independent set of utterances, where information from surrounding segments of speech is not utilised by the model. Of course, in many real-world scenarios, these segments of speech are highly interrelated, and the previously processed data can be used to form a useful prior on which to condition future transcription. While previous work has investigated the incorporation of context from surrounding utterances [2,3], doing so effectively with both linguistic and acoustic data remains an open problem. Indeed, many of the current architectures used for modelling sequential data are limited in their ability to utilise very long-distance dependencies [3]. This project aims to investigate approaches that enable automatic speech recognition systems to incorporate information from a much broader context. Ideally, in scenarios such as work-related meetings, ASR systems should be able to utilise relevant information from across the entire discourse to aid transcription. Additionally, an element of this work can look at what sorts of long-range dependencies are present in long-format speech datasets, and how well these are modelled by our current speech recognition systems. 1. Ebeling, Werner, and Thorsten Pöschel. "Entropy and long-range correlations in literary English." EPL (Europhysics Letters) 26, no. 4 (1994): 241. 2. G. Sun, C. Zhang and P. C. Woodland, "Transformer Language Models with LSTM-Based Cross-Utterance Information Representation" ICASSP (2021) 3. Hori, Takaaki, Niko Moritz, Chiori Hori and Jonathan Le Roux. "Advanced Long-context End-to-end Speech Recognition Using Context-expanded Transformers." Interspeech (2021). 4. Urvashi Khandelwal, He He, Peng Qi, and Dan Jurafsky. "Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context." ACL (2018)

  • Funder: UKRI Project Code: EP/K039237/1
    Funder Contribution: 1,075,140 GBP
    Partners: University of Sheffield

    Enhancing the safety of nuclear fuel is an important component in the continued use, and expansion, of nuclear power. One area where safety can be enhanced is enhancing the cladding around the nuclear fuel. Such a coating will enhance further the long term stability of the fuel under normal reactor operation, whilst at the same provide an extra level of insurance should an incident similar to that in Fukushima happen. These new coatings will provide a barrier between the Zircalloy cladding and air/water, which will help to prevent the formation of hydrogen gas from steam formed when there is a loss of coolant accident (LOCA), i.e. the process that happened at Fukushima Daichi, in March 2011. Using the combined expertise/knowledge from within the UK and US a collaborative research team has been put together to develop such coatings. Two options will be addressed one based on using oxide, such as zirconia, whilst a second will be based on ternary carbide/nitride based materials, such as MAX phases. M(n+1)AX(n) phases have previously been shown to not only recover rapidly from radiation damage, but also excellent thermal/corrosion properties making them ideal for this application. For the development of new coatings to be used in the current, and future, nuclear reactor fleet, new coatings must be prepared, and examined for stability, under a range of reactor conditions. The experimental programme will address issues such as the preparation of the coating, stability of bonding between coating and fuel, the effects of radiation damage on the interface, and how the enhanced coating increases stability of the fuel to both high temperatures/pressures experienced within a fission core. These experiments will also be used to validate simulations of corrosion, providing a means by which simulations can be reliably used. One final assessment of the coatings once tested, is how they behave under conditions that model a LOCA event. The results from this work will be used in developing technologies for existing and future reactor technologies.

  • Funder: EC Project Code: 303101
    Partners: University of Sheffield
  • Funder: UKRI Project Code: ES/G031770/1
    Funder Contribution: 63,163 GBP
    Partners: University of Sheffield

    Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.

  • Funder: UKRI Project Code: EP/F03363X/1
    Funder Contribution: 49,091 GBP
    Partners: University of Sheffield

    Recently, with the unfortunate emergence of bio-terrorism and its threat to both military targets and civilian populations, it is necessary to develop a portable and cheap system to continuously monitor for any potential aerosolized agents (biological particles) released from deadly biological weapons in any open area, even in harsh environments. As most bio-molecules show strong absorption in the ultra-violet (UV) spectral region ranging from 280 to 340 nm, an efficient UV lighting source is expected to be a crucial component for next-generation biological detection, biological imaging and disease analysis applications. In particular use of UV laser diodes would enable high sensitivity detection systems. III-nitride semiconductors are the best materials to make such laser diodes. In last decade, there have been major achievements in this area. However, the achievements are limited to the violet/blue spectral region, with those devices mainly based on the InGaN alloy. Due to a number of challenges in material growth, a 343 nm laser diode is the shortest one so far reported. Obviously, such a laser diode is not short enough to be employable for above applications.Target of this exploratory project is the development of the first 337 nm UV laser diode based on the GaN/AlGaN material system to replace currently used N2 gas-based lasers. This work is based on recent major advances of the here involved UK teams in the field of III-nitride semiconductors. Further applications of the technology involve biological imaging as an efficient method to detect diseases in a human body, for example, cancerous tissues.