Search

Personalised speech enhancement (PSE), which extracts only the speech of a target user and removes everything else from a recorded audio clip, can potentially improve users' experiences of audio AI modules deployed in the wild. To support a large variety of downstream audio tasks, such as real-time ASR and audio-call enhancement, a PSE solution should operate in a streaming mode, i.e., input audio cleaning should happen in real-time with a small latency and real-time factor. Personalisation is typically achieved by extracting a target speaker's voice profile from an enrolment audio, in the form of a static embedding vector, and then using it to condition the output of a PSE model. However, a fixed target speaker embedding may not be optimal under all conditions. In this work, we present a streaming Transformer-based PSE model and propose a novel cross-attention approach that gives adaptive target speaker representations. We present extensive experiments and show that our proposed cross-attention approach outperforms competitive baselines consistently, even when our model is only approximately half the size.

Blockchain technology, beyond cryptocurrencies, is called to be the new information exchange ecosystem due to its unique properties, such as immutability and transparency. The main objective of this work is to introduce the design of a decentralized rental system, which leverages smart contracts and the Ethereum public blockchain. The work started from an exhaustive investigation on the Ethereum platform, emphasizing the aspect of cryptography and all the technology behind this platform. In order to test the proposed scheme in a realistic use, the implementation of a web application for the rental of vehicles has been carried out. The application covers the entire vehicle rental process offered in traditional web applications, adding more autonomy and ease of use to users. Following Ethereum application development guidelines, all business logic is located in the smart contracts implemented in the Ethereum network, where these contracts control the entire vehicle rental system of customers. While this is a work in progress, the results obtained in the first proof of concept have been very promising.

Video surveillance systems are increasingly used in different fields, from the domestic to the commercial environment. Current systems are being improved and complemented with new elements and functionalities. This paper proposes the design of a video surveillance robot based on Raspberry Pi with the abilities to perform tasks of motion detection, send video on real time, fire detection and also, the possibility of control it remotely from the Internet. In order to check the information received from the robot, as well as the video sent, a client application has been developed to any device with an Internet connection. In addition to this, in order to protect the information obtained by the robot, a secure system is proposed, which uses different security mechanisms to achieve this goal.

This paper describes a solution for the efficient management of revocation in vehicular ad-hoc networks, for both certificate-based and identity-based authentication. It proposes the use of an authenticated data structure based on a dynamic hash tree, which is a perfect $k$-ary tree, together with a new version of the SHA-3 hash function. This combination allows optimizing search and insertion operations in the tree. Consequently, the proposal is very useful both when vehicular networks are widely used, and in urban environments. Simulation results are promising and confirm this hypothesis., Comment: arXiv admin note: substantial text overlap with arXiv:2208.05343

In some situations the communications of a place can be affected, totally lost, or not even exist. In these cases, the MANETs play an important role, allowing to establish a communications point using the different nodes of the network to reach the destination using decentralized communications. This paper proposes the implementation of a Robotic MANET, a decentralized network using robots as its nodes, which allows to move the network nodes to the desired location remotely. For this, each robot has as a core a Raspberry Pi with the capabilities to perform audio and video streaming, remote control of robots, tracking of objects, and deployment of wireless networks. To protect the network, different security mechanisms are used that allow secure authentication on the network by different nodes and encryption of information transmitted between them. All communications are protected through Identity-Based Cryptography, specifically with an Identity-Based Signcryption scheme.

Data mining and information extraction from data is a field that has gained relevance in recent years thanks to techniques based on artificial intelligence and use of machine and deep learning. The main aim of the present work is the development of a tool based on a previous behaviour study of security audit tools (oriented to SQL pentesting) with the purpose of creating testing sets capable of performing an accurate detection of a SQL attack. The study is based on the information collected through the generated web server logs in a pentesting laboratory environment. Then, making use of the common extracted patterns from the logs, each attack vector has been classified in risk levels (dangerous attack, normal attack, non-attack, etc.). Finally, a training with the generated data was performed in order to obtain a classifier system that has a variable performance between 97 and 99 percent in positive attack detection. The training data is shared to other servers in order to create a distributed network capable of deciding if a query is an attack or is a real petition and inform to connected clients in order to block the petitions from the attacker's IP.

A Mobile Ad hoc NETwork (MANET) is a type of wireless network without any infrastructure, where nodes must adapt to the changing dynamic situations that result from their mobility. Because of the decentralization of nodes and the security needs of communications, management of MANETs must be self-organized, which is a major research challenge. In order to cope with the intrinsic properties of MANETs, a new decentralized management system for MANETs called Self-organizing Life Cycle Management (SLCM), is here fully described and evaluated. Regarding security, node authentication is the most critical component of access control in any network, and in particular in MANETs. Broadcasting is also a fundamental data dissemination mechanism in these networks. Both aspects have received special attention when defining the proposed SLCM system. In particular, both a strong access control algorithm based on the cryptographic paradigm of zero-knowledge proofs, and a three-step broadcast protocol, are here defined. This work includes the performance evaluation of the scheme, and the obtained experimental results show that SLCM significantly improves both the quality and the security of life cycle management of self-organized MANETs., Comment: arXiv admin note: text overlap with arXiv:1005.3391, arXiv:1005.0106

Vehicular ad-hoc networks, where traffic information is distributed from many sources to many destinations, require data authentication mechanisms to detect any malicious behavior of users, such as modification or replay attacks. In this paper we propose a new data aggregation protocol that uses probabilistic verification to detect such attack attempts a posteriori in an efficient way, with minimal overhead and delay. The algorithm also contains an additional security mechanism based on reactive groups created on demand, which ensure a priori that vehicles generate trustworthy information. According to a comprehensive analysis including both a small-scale real device implementation and NS2 simulations, it is shown that the proposed protocol is robust.

A vehicular ad-hoc network is a type of communication network where the nodes are vehicles. Their aim is to manage traffic efficiently in order to prevent unwanted events such as accidents, jams, etc. The research on this type of networks assumes the use of additional infrastructure installed on the roads, and special devices installed in vehicles. In this paper we propose a novel solution for the deployment of vehicular networks, which uses only existing smartphones with technologies such as Wi-Fi, GPS and Bluetooth. This approach will save the costs of road infrastructure deployment and extra vehicles devices, making it possible to deploy the first real-world vehicular ad-hoc network with a low cost. In particular, the developed implementation of the proposal for the Symbian platform is here examined in full detail.

Neural networks (NNs) lack measures of "reliability" estimation that would enable reasoning over their predictions. Despite the vital importance, especially in areas of human well-being and health, state-of-the-art uncertainty estimation techniques are computationally expensive when applied to resource-constrained devices. We propose an efficient framework for predictive uncertainty estimation in NNs deployed on embedded edge systems with no need for fine-tuning or re-training strategies. To meet the energy and latency requirements of these embedded platforms the framework is built from the ground up to provide predictive uncertainty based only on one forward pass and a negligible amount of additional matrix multiplications with theoretically proven correctness. Our aim is to enable already trained deep learning models to generate uncertainty estimates on resource-limited devices at inference time focusing on classification tasks. This framework is founded on theoretical developments casting dropout training as approximate inference in Bayesian NNs. Our layerwise distribution approximation to the convolution layer cascades through the network, providing uncertainty estimates in one single run which ensures minimal overhead, especially compared with uncertainty techniques that require multiple forwards passes and an equal linear rise in energy and latency requirements making them unsuitable in practice. We demonstrate that it yields better performance and flexibility over previous work based on multilayer perceptrons to obtain uncertainty estimates. Our evaluation with mobile applications datasets shows that our approach not only obtains robust and accurate uncertainty estimations but also outperforms state-of-the-art methods in terms of systems performance, reducing energy consumption (up to 28x), keeping the memory overhead at a minimum while still improving accuracy (up to 16%)., Comment: 13 pages, 6 figures

LPCNet is an efficient vocoder that combines linear prediction and deep neural network modules to keep the computational complexity low. In this work, we present two techniques to further reduce it's complexity, aiming for a low-cost LPCNet vocoder-based neural Text-to-Speech (TTS) System. These techniques are: 1) Sample-bunching, which allows LPCNet to generate more than one audio sample per inference; and 2) Bit-bunching, which reduces the computations in the final layer of LPCNet. With the proposed bunching techniques, LPCNet, in conjunction with a Deep Convolutional TTS (DCTTS) acoustic model, shows a 2.19x improvement over the baseline run-time when running on a mobile device, with a less than 0.1 decrease in TTS mean opinion score (MOS)., Comment: Interspeech 2020

Increasing demand for on-device Automatic Speech Recognition (ASR) systems has resulted in renewed interests in developing automatic model compression techniques. Past research have shown that AutoML-based Low Rank Factorization (LRF) technique, when applied to an end-to-end Encoder-Attention-Decoder style ASR model, can achieve a speedup of up to 3.7x, outperforming laborious manual rank-selection approaches. However, we show that current AutoML-based search techniques only work up to a certain compression level, beyond which they fail to produce compressed models with acceptable word error rates (WER). In this work, we propose an iterative AutoML-based LRF approach that achieves over 5x compression without degrading the WER, thereby advancing the state-of-the-art in ASR compression.

The ability to form images through hair-thin optical fibres promises to open up new applications from biomedical imaging to industrial inspection. Unfortunately, deployment has been limited because small changes in mechanical deformation (e.g. bending) and temperature can completely scramble optical information, distorting images. Since such changes are dynamic, correcting them requires measurement of the fibre transmission matrix (TM) in situ immediately before imaging. TM calibration typically requires access to both the proximal and distal facets of the fibre simultaneously, which is not feasible during most realistic usage scenarios without compromising the thin form factor with bulky distal optics. Here, we introduce a new approach to determine the TM of multi-mode fibre (MMF) or multi-core fibre (MCF) in a reflection-mode configuration without access to the distal facet. A thin stack of structured metasurface reflectors is used at the distal facet to introduce wavelength-dependent, spatially heterogeneous reflectance profiles. We derive a first-order fibre model that compensates these wavelength-dependent changes in the TM and show that, consequently, the reflected data at 3 wavelengths can be used to unambiguously reconstruct the full TM by an iterative optimisation algorithm. We then present a method for sample illumination and imaging following TM reconstruction. Unlike previous approaches, our method does not require the TM to be unitary making it applicable to physically realistic fibre systems. We demonstrate TM reconstruction and imaging first using simulated non-unitary fibres and noisy reflection matrices, then using much larger experimentally-measured TMs of a densely-packed MCF, and finally on an experimentally-measured multi-wavelength set of TMs recorded from a MMF. Our findings pave the way for online transmission matrix calibration in situ in hair-thin optical fibres, Comment: Main text: 38 pages, 9 Figures, Appendices: 26 pages, 6 Figures. Corrected author affiliation

We introduce a framework for the reconstruction of the amplitude, phase and polarisation of an optical vector-field using calibration measurements acquired by an imaging device with an unknown linear transformation. By incorporating effective regularisation terms, this new approach is able to recover an optical vector-field with respect to an arbitrary representation system, which may be different from the one used in calibration. In particular, it enables the recovery of an optical vector-field with respect to a Fourier basis, which is shown to yield indicative features of increased scattering associated with tissue abnormalities. We demonstrate the effectiveness of our approach using synthetic holographic images as well as biological tissue samples in an experimental setting where measurements of an optical vector-field are acquired by a fibre endoscope, and observe that indeed the recovered Fourier coefficients are useful in distinguishing healthy tissues from lesions in early stages of oesophageal cancer.