A5007248971- Gaussian Processes and Bayesian Inference
- Control Systems and Identification
- Neurological disorders and treatments
- Advanced Multi-Objective Optimization Algorithms
- Fault Detection and Control Systems
- EEG and Brain-Computer Interfaces
- Advanced Neuroimaging Techniques and Applications
- Neural Networks and Applications
- Medical Image Segmentation Techniques
- Neuroscience and Neural Engineering
- Parkinson's Disease Mechanisms and Treatments
- Machine Learning and Data Classification
- Advanced MRI Techniques and Applications
- Face and Expression Recognition
- Blind Source Separation Techniques
- Tensor decomposition and applications
- Model Reduction and Neural Networks
- Time Series Analysis and Forecasting
- Emotion and Mood Recognition
- Speech and Audio Processing
- Video Coding and Compression Technologies
- Markov Chains and Monte Carlo Methods
- Image Retrieval and Classification Techniques
- Music and Audio Processing
- Speech Recognition and Synthesis
University of Manchester
2008-2025
University of Sheffield
2016-2024
Dalhousie University
2024
National Training Service
2018
Technological University of Pereira
2007-2017
Procter & Gamble (United States)
2016
Massachusetts Institute of Technology
2012
Universitat Politècnica de Catalunya
2002-2011
Max Planck Society
2010
Siemens (Spain)
2002
Recently there has been an increasing interest in regression methods that deal with multiple outputs. This motivated partly by frameworks like multitask learning, multisensor networks or structured output data. From a Gaussian processes perspective, the problem reduces to specifying appropriate covariance function that, whilst being positive semi-definite, captures dependencies between all data points and across One approach account for non-trivial correlations outputs employs convolution...
Purely data-driven approaches for machine learning present difficulties when data are scarce relative to the complexity of model or is forced extrapolate. On other hand, purely mechanistic need identify and specify all interactions in problem at hand (which may not be feasible) still leave issue how parameterize system. In this paper, we a hybrid approach using Gaussian processes differential equations combine modeling with physical We show different, physically inspired, kernel functions...
Multimodal Emotion recognition (MER) is an application of machine learning were different biological signals are used in order to automatically classify a determined affective state. MER systems has been developed for type applications from psychological evaluation, anxiety assessment, human-machine interfaces and marketing. There several spaces classification proposed the state art emotion task, most known discrete dimensional emotions described terms some basic latent dimensions...
This paper is concerned with learning and stochastic control in physical systems that contain unknown input signals. These signals are modeled as Gaussian processes (GP) certain parameterized covariance structures. The resulting latent force models can be seen hybrid a first-principle model part nonparametric GP part. We briefly review the statistical inference methods for this kind of models, introduce methodology these provide new theoretical observability controllability results them.
Abstract The key role of buildings in tackling climate change has gained global recognition. To avoid unnecessary costs and time wasted, it is important to understand the conditions energy usage for existing housing stock identify most features affecting consumption guide relevant retrofit measures. This paper investigated how spatial, morphological thermal characteristics residential houses contribute consumption. Additionally, presents a rapid assessment tool using minimum data input...
Motivation: Imaging Mass Cytometry (IMC) is a cutting-edge technology for analysing spatially resolved protein expression at the single-cell level. However, its downstream analyses are often hindered by batch effects, which introduce systematic biases and obscure true biological variations. Existing correction methods, largely developed scRNA-seq data, struggle to achieve precise control, leading either over-correction removing critical information, or under-correction leaving residual...
H.264/AVC is a new international video coding standard that provides higher efficiency with respect to previous standards at the expense of computational complexity. The complexity even when used in applications high bandwidth and quality like definition (HD) decoding. In this paper, we analyze requirements H.264 decoder special emphasis HD compare it lower resolutions. analysis was done SIMD optimized using hardware performance monitoring. main objective identify application bottlenecks...
Affective computing systems has a great potential in applications for biofeedback and cognitive conductual therapies. Here, by analyzing the physiological behavior of given subject, we can infer affective state an emotional process. Since, emotions be modeled as dynamic manifestations these signals, continuous analysis valence/arousal space, brings more information related to In this paper propose method affect recognition from multimodal signals. Our model is based on learning latent space...
Impacts of anthropogenic noise on human beings range from physiological to phycological and include direct damage auditory organs, elevated stress annoyance, impairments in mental development. Maritime transportation operations are also recognized as a major contributor nearby communities. The Port Halifax, Nova Scotia, Canada may be these concerns because its emitting it being near densely populated areas. This study examines the cost benefits implementing abatement measures reduce...
Emotion recognition is a challenging research problem with significant scientific interest. Most of the emotion assessment studies have focused on analysis facial expressions. Recently, it has been shown that simultaneous use several biosignals taken from patient may improve classification accuracy. An open in this area to identify which are more relevant for recognition. In paper, we perform Recursive Feature Elimination (RFE) select subset features allows classification. Experiments...