Noise measurement

About: Noise measurement is a research topic. Over the lifetime, 19776 publications have been published within this topic receiving 308180 citations.

Cepstral gain normalization for noise robust speech recognition

Abstract: The paper describes a robust speech recognition technique which normalizes cepstral gains in order to remove effects of additive noise. We assume that the effects can be expressed by an approximate model which consists of gain and DC components in log-spectrum. Accordingly, we propose cepstral gain normalization (CGN) which normalizes the gains by means of calculating maximum and minimum values of cepstral coefficients in speech frames. The proposed method can extract noise robust features without a priori knowledge and environmental adaptation because it is applied to both training and testing data. We have evaluated recognition performance under noisy environments using the Noisex-92 database and a 100 Japanese city names task. The CGN provides improvements of recognition accuracy at various SNRs compared with combinations of conventional methods.

Improving the Traceable Measurement and Generation of Small Direct Currents

Abstract: We present the latest improvements in the traceable measurement and generation of small electric currents A central tool in our traceability chain for small direct currents is a new binary cryogenic current comparator (CCC) with a total of 18 276 turns This 14-bit CCC is well suited for the calibration of high-value resistors and current amplifiers, but also for the direct amplification of small currents A noise level of 5 fA/ $\surd $ Hz at 005 Hz is routinely achieved The systematic uncertainty due to noise rectification was exemplarily investigated in a ratio-error test configuration, showing that a total uncertainty of about one part in $10^{6}$ can be achieved at 100 pA For further improvement, a new instrument was developed, the ultrastable low-noise current amplifier (ULCA) Its transfer coefficient is highly stable versus time, temperature, and current amplitude within a full dynamic range of ±5 nA The ULCA is calibrated with the 14-bit CCC at high current amplitude, and allows the measurement or generation of 100-pA direct current with an uncertainty of one part in $10^{7}$ The novel setup was successfully used to investigate the uncertainty of the established capacitor charging method A quantum metrology triangle experiment based on the presented instruments is proposed

Identifying learners robust to low quality data

Abstract: Real world datasets commonly contain noise that is distributed in both the independent and dependent variables. Noise, which typically consists of erroneous variable values, has been shown to significantly affect the classification performance of learners. In this study, we identify learners with robust performance in the presence of low quality (noisy) measurement data. Noise was injected into five class imbalanced software engineering measurement datasets, initially relatively free of noise. The experimental factors considered included the learner used, the level of injected noise, the dataset used (each with unique properties), and the percentage of minority instances containing noise. No other related studies were found that have identified learners that are robust in the presence of low quality measurement data. Based on the results of this study, we recommend using the random forest learner for building classification models from noisy data.

A Robust Scheme for Feature-Preserving Mesh Denoising

Abstract: In recent years researchers have made noticeable progresses in mesh denoising, that is, recovering high-quality 3D models from meshes corrupted with noise (raw or synthetic). Nevertheless, these state of the art approaches still fall short for robustly handling various noisy 3D models. The main technical challenge of robust mesh denoising is to remove noise while maximally preserving geometric features. In particular, this issue becomes more difficult for models with considerable amount of noise. In this paper we present a novel scheme for robust feature-preserving mesh denoising. Given a noisy mesh input, our method first estimates an initial mesh, then performs feature detection, identification and connection, and finally, iteratively updates vertex positions based on the constructed feature edges. Through many experiments, we show that our approach can robustly and effectively denoise various input mesh models with synthetic noise or raw scanned noise. The qualitative and quantitative comparisons between our method and the selected state of the art methods also show that our approach can noticeably outperform them in terms of both quality and robustness.

Non Adaptive Second-Order Generalized Integrator for Identification of a Biased Sinusoidal Signal

Abstract: This note presents a new algorithm that is designed to identify the frequency, magnitude, phase and offset of a biased sinusoidal signal. The structure of the algorithm includes an orthogonal system generator based on a second-order generalized integrator. The proposed strategy has the advantages of a fast and accurate signal reconstruction capability and a good rejection to noise.