The broadcasting industry has adopted IP technologies, revolutionising both live and pre-recorded content production, from news gathering to live music events. IP broadcasting allows for the transport of audio and video signals in an easily configurable way, aligning with modern networking techniques. This shift towards an IP workflow allows for much greater flexibility, not only in routing signals but with the integration of tools using standard web development techniques. One possible tool could include the use of live audio tagging, which has a number of uses in the production of content. These could include adding sound effects to automated closed captioning or identifying unwanted sound events within a scene. In this paper, we describe the process of containerising an audio tagging model into a microservice, a small segregated code module that can be integrated into a multitude of different network setups. The goal is to develop a modular, accessible, and flexible tool capable of seamless deployment into broadcasting workflows of all sizes, from small productions to large corporations. Challenges surrounding latency of the selected audio tagging model and its effect on the usefulness of the end product are discussed.
University of Surrey - Research Portal
Welcome to the University of Surrey Open Research repository
The University of Surrey is a world-class, research-led university committed to research excellence and greater access to research for all. Our research seeks to answer global challenges, drive innovation and deliver real-world impact.
Discover here the fantastic research produced by the University’s staff and postgraduate research students.
Downloads by country since 11th November 2020
Country | Downloads | Percent |
---|---|---|
United States | 1233632 | 12% |
China | 1034771 | 10% |
United Kingdom | 342979 | 3.36% |
Singapore | 121247 | 1.19% |
Germany | 111993 | 1.1% |
All others | 7360470 | 72% |
Repository metrics
Top Ten
Conference paper
Integrating IP broadcasting with audio tags: Workflow and challenges
by Rhys Burchett-Vass, Arshdeep Singh, Gabriel Bibbó and Mark D. Plumbley
Book
Beyond Porridge: Dishes from Inside Women’s Prisons
by Maria Adams, Jon Garland, Vicki Harman, Daniel McCarthy, Erin Power and Talitha Brown
An illustrated collection of recipes and advice based around the experiences of women separated from their families by imprisonment. Beyond Porridge engages with creativity, social interactions and diversity within prison walls. The book explains how women in prison use limited ingredients to supplement their standard prison diet and enjoy a 'minor feast'. It tells of the comfort foods they miss most. In their own words and illustrations the book shows how ingenuity can spring from the most unlikely of circumstances.The book is based on a project led by researchers at the University of Surrey. It introduces readers to Chocolate Pudding in a Mug, a quick version of Callaloo prepared in a prison cell, Mackerel Curry, Pumpkin Seed Fry, Oreo Cheesecake and Kettle Chow Mein.
Journal article
by Ellie E. Miles, Jennifer L. Nicol, Hatti Fowler, Amelia Roberts, Andrew T. Hulton, Caitlin Jeary, Renata Walewska, Sunil Iyengar, Erik D. Hanson and Andrea Sitlinger ... (12 authors)
Background
Chronic lymphocytic leukaemia (CLL) has a heterogeneous lifelong course. While some patients never require treatment, most experience intermittent periods of active monitoring with other time points in active treatment. Most patients experience significant symptoms which negatively impact their quality of life (QoL). Although physical activity and exercise may help manage symptoms, it is unclear what disease-related factors drive the physical inactivity observed in people with CLL.
Methods
This study explored physical activity among people with CLL and assessed differences and relationships in treatment stage, symptoms, quality of life, and preferences for physical activity using an online questionnaire.
Results
A total of 128 individuals with CLL [66 M/62F: mean age 67 ± 9.1 years (range 38–91 years)] completed the questionnaire. Those who are being/have been treated (N = 55) exhibited worse QoL (p = 0.018) and lower engagement in higher levels of physical activity (p = 0.045) when compared to their treatment naïve (N = 73) counterparts. Both groups had similar symptomology, with fatigue (∼77%) and insomnia (∼55%) being the most reported and associated with less likelihood of being physically active. Physically active participants reported better QoL (p = 0.020), physical functioning (p = 0.003) and role functioning (p = 0.020) as well as lower levels of fatigue (p = 0.036), pain (p = 0.017) and symptom burden (p = 0.026) compared to those who were insufficiently active. Although 79% of respondents wanted to engage in exercise programs for their CLL, 70% reported never receiving exercise guidance from their healthcare professionals.
Conclusion
These findings highlight a significant need for targeted interventions to increase physical activity, likely improving QoL, in people with CLL. Furthermore, there is considerable interest from the CLL community in receiving exercise guidance; however, factors such as treatment status and symptomology should be considered when developing CLL-specific exercise programs.
Journal article
A Hybrid Framework for Soil Property Estimation from Hyperspectral Imaging
by Daniel L Ayuba, Jean-Yves Guillemaut, Belen Marti-Cardona and Oscar Mendez
Accurate estimation of soil properties is crucial for optimizing agricultural practices and promoting sustainable resource management. Hyperspectral imaging provides a non-invasive means of quantifying key soil parameters, but effectively utilizing the high-dimensional hyperspectral data presents significant challenges. In this paper, we introduce HyperSoilNet, a hybrid deep learning framework for estimating soil properties from hyperspectral imagery. HyperSoilNet leverages a pretrained hyperspectral-native CNN backbone and integrates it with a carefully optimized machine learning (ML) ensemble to combine the strengths of deep representation learning with traditional ML techniques. We evaluate our framework on the Hyperview challenge dataset, focusing on four critical soil properties: potassium oxide, phosphorus pentoxide, magnesium, and soil pH. Comprehensive experiments demonstrate that HyperSoilNet surpasses state-of-the-art models, achieving a score of 0.762 on the challenge leaderboard. Through detailed ablation studies and spectral analysis, we provide insights on the components of the framework, and their contribution to performance, showcasing its potential for advancing precision agriculture and sustainable soil management practices.
Journal article
by Ilaria Riboldi, Cristina Crocamo, Chiara Alessandra Capogrosso, Francesco Bartoli, Jo Armes, Cath Taylor and Giuseppe Carrà
Background/Objectives: Both traumatic and stressful events, including major life changes, may contribute to post-traumatic stress symptoms (PTS), often associated with anxiety and depression. Feelings of loneliness may influence these relationships, whilst social support seems to mitigate the effects of stressful events on mental health. Our study thus aimed to evaluate the mediating role of loneliness in the relationships between PTS and both anxiety and depressive symptoms among university students. Methods: The data were from the CAMPUS study (0058642/21; FHMS 20-21 157), a survey on university students’ mental health in Italy and the UK. Using a logit model, mediation analyses were carried out to test whether the relationships between PTS and both anxiety and depressive symptoms might be mediated by loneliness. A path analysis was then performed to jointly test the associations between the Impact of Event Scale—Revised (IES-R)’s subscales and clinical domains. Results: Positive associations were found between PTS and both anxiety (p < 0.001) and depressive symptoms (p < 0.001). However, loneliness mediated approximately 22% of the effect of the PTS on anxiety symptoms (indirect effect: 1.04, 95% CI: 0.59; 1.48, p < 0.001) and approximately 33% of the effect of the PTS on depressive symptoms (indirect effect: 1.81, 95% CI: 1.22; 2.39, p < 0.001). Furthermore, the path analysis indicated associations between the IES-R’s hyperarousal subscale and both anxiety (coeff.: 0.34, p < 0.001) and depressive symptoms (coeff.: 0.27, p < 0.001). Conclusions: Along with the associations between PTS and both anxiety and depressive symptoms, our findings highlight the key role of loneliness in both these associations. Targeted interventions to reduce loneliness, especially for students exposed to traumatic events, may ultimately improve their mental health.
Book chapter
by Lena Mattheis
Nonbinary narratives are not just defined by genderqueer narrators or focalisers; they also move beyond binaries in the way they shape form and language. The study of nonbinary narratives therefore necessitates nonbinary thought and methods. In this essay, I therefore identify some of the ways in which texts can be read and analysed in a nonbinary manner. Building on queer, feminist and trans narrative studies, and on Kanaka Maoli and Two-Spirit understandings of what constitutes binary-defying narratives, I present an overview of current nonbinary narrative research. In order to illustrate what a nonbinary approach to narration might look like, I present three brief case studies of the use of singular ‘they’ in Rae Spoon’s novel Green Glass Ghosts (2021), the nonbinary form of Sara Taylor's The Lauras (2017) and genre transgressions in Joshua Whitehead's Making Love with the Land (2022).
Conference proceeding
VisualSpeaker: Visually-Guided 3D Avatar Lip Synthesis
by Alexandre Symeonidis-Herzig, Ozge Mercanoglu Sincan and Richard Bowden
Realistic, high-fidelity 3D facial animations are crucial for expressive avatar systems in human-computer interaction and accessibility. Although prior methods show promising quality, their reliance on the mesh domain limits their ability to fully leverage the rapid visual innovations seen in 2D computer vision and graphics. We propose VisualSpeaker, a novel method that bridges this gap using photorealistic differentiable rendering, supervised by visual speech recognition, for improved 3D facial animation. Our contribution is a perceptual lip-reading loss, derived by passing photorealistic 3D Gaussian Splatting avatar renders through a pre-trained Visual Automatic Speech Recognition model during training. Evaluation on the MEAD dataset demonstrates that VisualSpeaker improves both the standard Lip Vertex Error metric by 56.1% and the perceptual quality of the generated animations, while retaining the controllability of mesh-driven animation. This perceptual focus naturally supports accurate mouthings, essential cues that disambiguate similar manual signs in sign language avatars.
Conference proceeding
SAGE: Segment-Aware Gloss-Free Encoding for Token-Efficient Sign Language Translation
by Jian He Low, Ozge Mercanoglu Sincan and Richard Bowden Prof
Gloss-free Sign Language Translation (SLT) has advanced rapidly, achieving strong performances without relying on gloss annotations. However, these gains have often come with increased model complexity and high computational demands, raising concerns about scalability, especially as large-scale sign language datasets become more common. We propose a segment-aware visual tokenization framework that leverages sign segmentation to convert continuous video into discrete, sign-informed visual tokens. This reduces input sequence length by up to 50% compared to prior methods, resulting in up to 2.67× lower memory usage and better scalability on larger datasets. To bridge the visual and linguistic modalities, we introduce a token-to-token contrastive alignment objective, along with a dual-level supervision that aligns both language embeddings and intermediate hidden states. This improves fine-grained cross-modal alignment without relying on gloss-level supervision. Our approach notably exceeds the performance of state-of-the-art methods on the PHOENIX14T benchmark, while significantly reducing sequence length. Further experiments also demonstrate our improved performance over prior work under comparable sequence-lengths, validating the potential of our tokenization and alignment strategies.
Conference proceeding
4D FMCW MIMO radar based MCMC-EPF Track-Before-Detect method for UAV tracking in low SNR
by Yingquan Zou, Jiayu Peng, Jingfu Li, Chong Huang and Pei Xiao
Tracking micro-unmanned aerial vehicles (micro-UAVs) in low signal-to-noise ratio (SNR) environments poses significant challenges due to their weak radar cross-section (RCS) and the inherent limitations of traditional Detect-Before-Track (DBT) radar algorithms. This paper proposes a novel Track-Before-Detect (TBD) approach based on a Markov Chain Monte Carlo-Enhanced Particle Filter (MCMC-EPF), leveraging 4D Frequency Modulated Continuous Wave (FMCW) MIMO radar. By directly processing unthresholded multi-frame 4D-FFT radar data, the method achieves joint detection and tracking, effectively preserving weak target information that is typically lost in DBT methods. Experimental results on real radar data demonstrate that the proposed algorithm achieves robust and accurate UAV tracking, maintaining a root-mean-square error (RMSE) within 1 meter and an average relative tracking error below 2.5% under low SNR conditions. These results highlight the method's potential for reliable UAV surveillance in challenging operational environments.
Report
New approaches to women-centred food policy and practice in prison: A toolkit for women’s prisons
by Maria Adams, Jon Garland, Vicki Harman, Daniel McCarthy, Erin Power and Talitha Brown
This toolkit is designed for prison governors and staff to improve the quality of food. It will help those involved in the preparation and production of food in women’s prisons to reflect on existing practices and to identify ways forward to improve the relationship between women and food in prison. The material in the toolkit is drawn from ESRC- and British Academy/Leverhulme-funded qualitative studies that focused on the role of food in women’s prisons. It has been devised in partnership with His Majesty’s Prison and Probation Service and charitable organisations including Women in Prison, Food Behind Bars, Food Matters and PACT.