This paper presents a new approach for supervised power disaggregation by using deep recurrent long short term memory network. It is useful to extract the signal of one dominant appliance or any subcircuit from aggregate signal. To train network, measurement target in addition total during same time period required. The method supervised, but less restrictive practice since submetering an important feasible. main advantages this are: a) also applicable variable load and not restricted on-off...

In this work, we further develop the conformerbased metric generative adversarial network (CMGAN) model <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> for speech enhancement (SE) in time-frequency (TF) domain. This paper builds on our previous work but takes a more indepth look by conducting extensive ablation studies inputs and architectural design choices. We rigorously tested generalization ability of to unseen noise types...

Experts assume that accidents caused by drowsiness are significantly under-reported in police crash investigations (1-3%). They estimate about 24-33% of the severe related to drowsiness. In order develop warning systems detect reduced vigilance based on driving behavior, a reliable and accurate reference is needed. Studies have shown measures driver's eyes capable under simulator or experiment conditions. this study, performance latest eye tracking in-vehicle fatigue prediction evaluated....

In this paper, a radar demonstrator system with real-time capability operating at W -band is presented. It operates 90-100 GHz and provides 3-D information about the illuminated scene. The uses frequency modulated continuous wave signals to extract range whereupon long-range applications are aimed at. consists of sparse array 22 transmitting receiving antennas makes use multiple input output (MIMO) principle. A back-propagation algorithm cross-range information. With help simulations...

Purpose Motion is 1 extrinsic source for imaging artifacts in MRI that can strongly deteriorate image quality and, thus, impair diagnostic accuracy. In addition to involuntary physiological motion such as respiration and cardiac motion, intended accidental patient movements occur. Any impairment by reduce the reliability precision of diagnosis a motion‐free reacquisition become time‐ cost‐intensive. Numerous correction strategies have been proposed or prevent artifacts. These methods common...

Scene understanding for automated driving requires accurate detection and classification of objects other traffic participants. Automotive radar has shown great potential as a sensor driver assistance systems due to its robustness weather light conditions, but reliable object types in real time proved be very challenging. Here we propose novel concept radar-based classification, which utilizes the power modern Deep Learning methods learn favorable data representations thereby replaces large...

Image-to-image translation is a new field in computer vision with multiple potential applications the medical domain. However, for supervised image frameworks, co-registered datasets, paired pixel-wise sense, are required. This often difficult to acquire realistic scenarios. On other hand, unsupervised frameworks result blurred translated images unrealistic details. In this work, we propose framework which titled Cycle-MedGAN. The proposed utilizes non-adversarial cycle losses direct...

Since distribution shifts are likely to occur during testtime and can drastically decrease the model's performance, online test-time adaptation (TTA) continues update model after deployment, leveraging current test data. Clearly, a method proposed for TTA has perform well all kinds of environmental conditions. By introducing variable factors domain non-stationarity temporal correlation, we first unfold practically relevant settings define entity as universal TTA. We want highlight that this...

This paper presents a theoretical analysis of the Cramer-Rao lower bound for source localization from time differences arrival. We derive properties and design optimum sensor arrays which minimize bound.

This paper investigates the classification of different emotional states using presodic and voice quality information. We want to exploit usage phonation types within production emotions. Therefore, as features we use prosodic features, parameters, combinations both types. study how overlap or complement each other in application emotion recognition. The is speaker independent uses a reduced subset 8 Bayesian classifier.

Abstract Purpose: To obtain quantitative measures of human body fat compartments from whole MR datasets for the risk estimation in subjects prone to metabolic diseases without need any user interaction or expert knowledge. Materials and Methods: Sets axial T1‐weighted spin‐echo images were acquired. The segmented using a modified fuzzy c‐means algorithm. A separation into anatomic regions along axis was performed define with visceral adipose tissue present, standardize results. In abdominal...

Nowadays, renewable energies play an important role to cover the increasing power demand in accordance with environment protection. Solar energy, produced by large solar farms, is a fast growing technology offering environmental friendly supply. However, its efficiency suffers from cell defects occurring during operation life or caused incidents. These can be made visible using electroluminescence (EL) imaging. A manual classification of these EL images very time and cost demanding prone...

A fundamental drawback of the orthogonal frequency division multiplexing (OFDM), both in communication and radar, is its sensitivity to Doppler effect. In this paper, we present a novel signal processing approach for OFDM radar systems that overcomes OFDM. We propose scenario independent correction method enables an intercarrier-interference free The robustness proposed opens new perspectives system parameterization, enabling concepts not feasible before. show simulations superior classical...

Abstract Background Attenuation correction (AC) of PET data is usually performed using a second imaging for the generation attenuation maps. In certain situations however—when CT- or MR-derived maps are corrupted CT acquisition solely purpose AC shall be avoided—it would value to have possibility obtaining only based on information. The this study was thus develop, implement, and evaluate deep learning-based method whole body [ 18 F]FDG-PET which independent other modalities acquiring map....

In this letter, we present an extension of the PASTd algorithm to both rank and signal subspace tracking. It has a low computational complexity O(nr), where n is input vector length, r denotes dimension. Its performance in tracking time-varying direction arrival comparable with that expensive eigenvalue decomposition more robust than O(n/sup 2/) revealing URV updating proposed by Stewart.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

This paper presents a new supervised approach to extract the power trace of individual loads from single channel aggregate signals in non-intrusive load monitoring (NILM) systems. Recent approaches this source separation problem are based on factorial hidden Markov models (FHMM). Drawbacks needed knowledge HMM for all loads, what is infeasible large buildings, and combinatorial complexity. Our trains with two emission probabilities, one be extracted other signal. A Gaussian distribution used...

Event detection plays an important role in today's Non-Intrusive Load Monitoring (NILM) systems faced more and with nonlinear variable loads. For this purpose, the paper presents unsupervised NILM event detector based on kernel Fisher discriminant analysis (KFDA) which provides accurate start end times of so-called active sections. Active sections are extension classical events introduced to include pulses, load intervals noisy signals makes classification flexible. The achieves good...

Numerous factors could lead to partial deteriorations of medical images. For example, metallic implants will localized perturbations in MRI scans. This affect further post-processing tasks such as attenuation correction PET/MRI or radiation therapy planning. In this work, we propose the inpainting images via Generative Adversarial Networks (GANs). The proposed framework incorporates two patch-based discriminator networks with additional style and perceptual losses for missing information...

The understanding of the surrounding environment plays a critical role in autonomous robotic systems, such as self-driving cars. Extensive research has been carried out concerning visual perception. Yet, to obtain more complete perception environment, systems future should also take acoustic information into account. Recent sound event localization and detection (SELD) frameworks utilize convolutional recurrent neural networks (CRNNs). However, considering nature CRNNs, it becomes...

A Cartesian subsampling scheme is proposed incorporating the idea of PF acquisition and variable-density Poisson Disc (vdPD) by redistributing sampling space onto a smaller region aiming to increase k-space density for given acceleration factor. Especially normally sparse sampled high-frequency components benefit from this redistribution, leading improved edge delineation. The prospective subsampled compacted can be reconstructed seamless combination CS-algorithm with Hermitian symmetry...