Sorry, you need to enable JavaScript to visit this website.

Many image-based rendering (IBR) methods rely on depth estimates obtained from structured light or time-of-flight depth sensors to synthesize novel views from sparse camera networks. However, these estimates often contain missing or noisy regions, resulting in an incorrect mapping between source and target views. This situation makes the fusion process more challenging, as the visual information is misaligned, inconsistent, or missing.

Previous research on the diagnosis of Bipolar disorder has mainly focused on resting-state functional magnetic resonance imaging. However, their accuracy can not meet the requirements of clinical diagnosis. Efficient multimodal fusion strategies have great potential for applications in multimodal data and can further improve the performance of medical diagnosis models. In this work, we utilize both sMRI and fMRI data and propose a novel multimodal diagnosis model for bipolar disorder.

Quasi-continuous system identification of time-variant linear acoustic systems can be applied in various audio signal processing applications when numerous acoustic transfer functions must be measured. A prominent application is measuring head-related transfer functions. We treat the underlying multiple-input-multiple-output (MIMO) system identification problem in a state-space model as a joint estimation problem for states, representing impulse responses, and state-space model parameters using the expectation maximization (EM) algorithm.

Clusters of neurons generate electrical signals which propagate in all directions through brain tissue, skull, and scalp of different conductivity. Measuring these signals with electroencephalography (EEG) sensors placed on the scalp results in noisy data. This can have severe impact on estimation, such as, source localization and temporal response functions (TRFs). We hypothesize that some of the noise is due to a Wiener-structured signal propagation with both linear and nonlinear components.

This paper proposes a reversible jump Markov chain Monte Carlo method that provides efficient inference for the general problem of pulse fitting. In particular, it minimises the potential of an adopted parametric model overfitting to the (noisy) data via the inclusion of a peak proximity parameter. This facilitates learning a more representative underlying model and significantly reduces the computational cost. Synthetic and real data are used to demonstrate the efficacy of the introduced Bayesian technique.

Contact: afg30@cam.ac.uk

Existing human object interaction (HOI) detection methods have introduced zero-shot learning techniques to recognize unseen interactions, but they still have limitations in understanding context information and comprehensive reasoning. To overcome these limitations, we propose a novel HOI learning framework, ContextHOI, which serves as an effective contextual HOI detector to enhance contextual understanding and zero-shot reasoning ability. The main contributions of the proposed ContextHOI are a novel context-mining decoder and a powerful interaction reasoning large language model (LLM).

Protein representation learning is critical in various tasks in biology, such as drug design and protein structure or function prediction, which has primarily benefited from protein language models and graph neural networks. These models can capture intrinsic patterns from protein sequences and structures through masking and task-related losses. However, the learned protein representations are usually not well optimized, leading to performance degradation due to limited data, difficulty adapting to new tasks, etc.

Neural Radiance Fields (NeRF) have revolutionized 3D scene modeling and rendering. However, their performance dips when handling images with diverse exposure levels, mainly due to the intricate luminance dynamics. Addressing this, we present an innovative method that proficiently models and renders images across a spectrum of exposure conditions. Our approach utilizes an unsupervised classifier-generator structure for HDR fusion, significantly enhancing NeRF's ability to comprehend and adjust to light variations, leading to the generation of images with appropriate brightness.

End-to-end paradigm has aroused more and more interests and attention for improving speech-to-text translation (ST) recently. Existing end-to-end models mainly attributes and attempts to address the problem of modeling burden and data scarcity, while always fail to maintain both cross-modal and cross-lingual mapping well at the same time.

This paper works on non-autoregressive automatic speech recognition. A unimodal aggregation (UMA) is proposed to segment and integrate the feature frames that belong to the same text token, and thus to learn better feature representations for text tokens. The frame-wise features and weights are both derived from an encoder. Then, the feature frames with unimodal weights are integrated and further processed by a decoder. Connectionist temporal classification (CTC) loss is applied for training.