28 accepted papers
This letter presents a modular micro-UAV system based on electro-permanent magnet (EPM) technology, addressing the critical challenges of energy-constrained docking mechanisms in resource-limited micro aerial platforms. Built upon the Crazyflie 2.1 platform, our 75 g modular design (with battery) fe
Tactile sensing is essential for robotic perception in scenarios where visual input is limited or unavailable. In this work, we propose a fully tactile-based 3D object reconstruction framework that recovers object shapes exclusively from contact observations. A robotic system comprising a robotic ar
Representation learning is fundamental to modern machine learning, powering applications such as text retrieval and multimodal understanding. However, learning robust and generalizable representations remains challenging. While prior work has demonstrated that active noise injection, a form of data
Recent advances have empowered large language models (LLMs) with remarkable fine-grained instruction-following capabilities in text generation tasks. However, embedding methods typically rely solely on the hidden state of the input's last token, limiting their ability to capture complete semantic si…
Intuitive human–robot interfaces are essential to increase usability and personalization in wearable robotic assistive technologies. However, most current systems rely on pre-programmed or sensor-driven strategies that offer limited active user control online. To address this limitation, we present …
Partially deepfake audio localization is important in audio forensics. However, existing localization models for partially deepfake audio face two major challenges: distribution shifts between training and testing data as well as insufficient utilization of information from both manipulated regions…
Recent works have shown that prompting large language models with audio encodings can unlock speech recognition capabilities. However, existing techniques do not scale efficiently, especially while handling long form streaming audio inputs — not only do they extrapolate poorly beyond the audio lengt…
We propose a multimodal 3D scene editing framework MMEditor to create or modify objects within an extant 3D Gaussian Splatting (3DGS) according to text and image prompts. MMEditor employs a multimodal image editing module to iteratively optimize 3D Gaussians in editing regions for delicate and multi…
Rapid advancements in speech synthesis and voice conversion bring convenience but also new security risks, creating an urgent need for effective audio deepfake detection. Although current models perform well, their effectiveness diminishes when confronted with the diverse and evolving nature of real…
The imitation of voice, targeted on specific speech attributes such as timbre and speaking style, is crucial in speech generation. However, existing methods rely heavily on annotated data, and struggle with effectively disentangling timbre and style, leading to challenges in achieving controllable g…
Multi-robot systems (MRSs) are becoming increasingly important in various domains. However, effective communication and coordination among multiple robots remain significant challenges. In this letter, we introduce a novel architecture for multi-robot decision-making and control based on multi-agent
Connectionist temporal classification (CTC) models are known to have peaky output distributions. Such behavior is not a problem for automatic speech recognition (ASR), but it can cause inaccurate forced alignments (FA), especially at finer granularity, e.g., phoneme level. This paper aims at allevia…
With the widespread application of Automatic Speaker Verification (ASV) technology in security authentication, the threat of fake audio attacks looms as a malicious means compromising system security. In this study, we employ the multi-scale permutation entropy (MPE) in audio deepfake detection, whi…
Multimodal learning which integrates data from diverse sensory modes plays a pivotal role in artificial intelligence. However existing multimodal learning methods often struggle with challenges where some modalities appear more dominant than others during multimodal learning resulting in suboptimal…
The rapid evolution of speech synthesis and voice conversion has raised substantial concerns due to the potential misuse of such technology, prompting a pressing need for effective audio deepfake detection mechanisms. Existing detection models have shown remarkable success in discriminating known de…
Neural transducers have achieved human level performance on standard speech recognition benchmarks. However, their performance significantly degrades in the presence of cross-talk, especially when the primary speaker has a low signal-to-noise ratio. Anchored speech recognition refers to a class of m…
Current fake audio detection algorithms have achieved promising performances on most datasets. However, their performance may be significantly degraded when dealing with audio of a different dataset. The orthogonal weight modification to overcome catastrophic forgetting does not consider the similar…
Just like in humans vision plays a fundamental role in guiding adaptive locomotion, when designing the control strategy for a walking assistive technology, the use of computer vision may substantially improve modulation of the assistance based on the external environment. In this letter, we develope
Measuring quality and intelligibility of a speech signal is usually a critical step in development of speech processing systems. To enable this, a variety of metrics to measure quality and intelligibility under different assumptions have been developed. Through this paper, we introduce tools and a s…
In the field of rehabilitation robotics, transparent, precise and intuitive control of hand exoskeletons still represents a substantial challenge. In particular, the use of compliant systems often leads to a trade-off between lightness and material flexibility, and control precision. In this letter,
Optimally synchronising the assistance provided by wearable devices with the human voluntary motion is still an open challenge in robotics. In order to provide accurate and robust assistance, this paper presents a novel approach that combines a layered implementation of a controller for an underactu
From wearables to powerful smart devices, modern automatic speech recognition (ASR) models run on a variety of edge devices with different computational budgets. To navigate the Pareto front of model accuracy vs model size, researchers are trapped in a dilemma of optimizing model accuracy by trainin…
This paper improves the streaming transformer transducer for speech recognition using non-causal convolution. Many works apply the causal convolution to improve streaming transformer ignoring the lookahead context. We propose to use non-causal convolution to process the center block and lookahead co…
The problem of machine learning systems demonstrating bias towards specific groups of individuals has been studied extensively, particularly in the Facial Recognition area, but much less so in Automatic Speech Recognition (ASR). This paper presents initial Speech Recognition results on “Casual Conve…
Reproducing the mechanisms of human locomotion is a hard challenge. Assistive wearable devices in this context need to be lightweight, portable, and to adapt to the wearer’s walking pattern. Aiming to combine the aforementioned features, we developed a soft wearable exosuit to assist hip flexion dur
Deep acoustic models typically receive features in the first layer of the network, and process increasingly abstract representations in the subsequent layers. Here, we propose to feed the input features at multiple depths in the acoustic model. As our motivation is to allow acoustic models to re-exa…
In this paper, we investigate out-of-vocabulary (OOV) word recovery in hybrid automatic speech recognition (ASR) systems, with emphasis on dynamic vocabulary expansion for both Weight Finite State Transducer (WFST)-based decoding and word-level RNNLM rescoring. We first describe our OOV candidate ge…
We propose and evaluate transformer-based acoustic models (AMs) for hybrid speech recognition. Several modeling choices are discussed in this work, including various positional embedding methods and an iterated loss to enable training deep transformers. We also present a preliminary study of using l…