| 1. |
- Akesson, Maria, et al.
(författare)
-
Highly Accurate and Noise-Robust Phase Delay Estimation using Multitaper Reassignment
- 2023
-
Ingår i: 31st European Signal Processing Conference, EUSIPCO 2023 - Proceedings. - 2219-5491. - 9789464593600 ; , s. 1763-1767
-
Konferensbidrag (refereegranskat)abstract
- The recently developed Phase-Scaled Reassignment (PSR) can estimate phase-difference between two oscillating transient signals with high accuracy. However, in low signal-to-noise ratios (SNRs) the performance of commonly applied reassignment techniques is known to deteriorate. In order to reduce variance in low SNR, we propose a multitaper PSR (mtPSR) method for phase-difference estimation between Gaussian transient signals. Three possible variations of this method are investigated and evaluated, mtPSR1, mtPSR2, and mtPSR3. All three variations are shown to outperform state-of-the-art methods and improve estimation accuracy in low SNR. The mtPSR1 is superior in terms of computational efficiency while the mtPSR3 achieves the highest accuracy. The mtPSR technique is also shown to be robust to model assumptions. An example of phase delay estimates of the electrical signals measured from the brain reveals promising results.
|
|
| 2. |
- Brynolfsson, Johan, et al.
(författare)
-
A time-frequency-shift invariant parameter estimator for oscillating transient functions using the matched window reassignment
- 2021
-
Ingår i: Signal Processing. - : Elsevier BV. - 0165-1684. ; 183
-
Tidskriftsartikel (refereegranskat)abstract
- In this paper we present the matched window reassignment method, generalizing the results to complex valued signals in multiple dimensions. For an oscillating transient signal with an envelope shape described by an arbitrary twice differentiable function, the reassigned spectrogram, with a matched window, concentrates all energy into one single time-frequency point. An estimator for the parameters of the envelope, in multiple dimensions, is constructed using the above property where the concentration is measured using the Rényi entropy. Furthermore, we present a classification scheme, where an observation is classified based on the concentration when reassigning with a set of model functions. Finally, two examples of parameter estimation from real-world measurements are shown, a one-dimensional time series of a single dolphin click and a two-dimensional time-series of seismic data.
|
|
| 3. |
- Reinhold, Isabella, et al.
(författare)
-
A novel Doppler penalty function for the multitaper Wigner-Ville distribution
- 2016
-
Ingår i: Proceedings of the 11th IMA International Conference on Mathematics in Signal Processing.
-
Konferensbidrag (refereegranskat)abstract
- This paper proposes a novel Doppler penalty kernel used in the computation of multitapers corresponding to the Wigner-Ville distribution. The resulting multitaper spectrogram will approximately fulfil the time-frequency concentration of the Wigner-Ville distribution but also suppress the cross-terms outside a signal dependent Doppler bandwidth, making the penalty kernel appropriate for linear frequency modulated signals. A limited number of tapers are used which results in computationally efficient calculations. A method for selecting a signal dependent Doppler bandwidth which balances the trade-off between auto-term resolution and cross-term suppression is also proposed. The time-frequency concentration of the proposed method is compared to other penalty kernel methods and a lag-independent kernel with compact support. The performance for white noise disturbances is also evaluated. Estimation of heart rate variability data exemplify the use of the penalty kernel.
|
|
| 4. |
- Reinhold, Isabella, et al.
(författare)
-
Efficient Thomson Spectral Estimator with Time-shifted Windows
- 2019
-
Ingår i: 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. - 9781479981311 ; 2019-May, s. 4983-4987
-
Konferensbidrag (refereegranskat)abstract
- In this paper optimal spectral analysis window shapes, using weighted discrete prolate spheroidal sequences as basis functions, are proposed. These windows are not typically positive or even. The windows are time-shifted, combining the computational efficiency of the Welch method and the appealing property of predefined frequency resolution of the Thomson spectral estimator. The parameters of the optimal windows are found by minimising the resulting spectral covariances and optimising the window overlap, for the predetermined frequency resolution and number of windows. The windows are found to have low side lobes, giving small spectral leakage, and the final spectral estimate gives close to optimal variance reduction, i.e. the covariance between different sub-spectra is very small.
|
|
| 5. |
- Reinhold, Isabella
(författare)
-
High resolution time-frequency representations
- 2017
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Non-stationary signals are very common in nature, e.g. sound waves such as human speech, bird song and music. It is usually meaningful to describe a signal in terms of time and frequency. Methods for doing so exist and are well defined. From the time representation it is possible to see the oscillations or waves of the signal and if the signal changes over time. From the frequency representation, obtained from the Fourier transform, the frequency decomposition of the signal can be seen, i.e. which frequencies the signal contains.However the time and frequency representations are not unique for any given signal, i.e. the transformation from time to frequency is not injective. It is therefore, especially for non-stationary and multi-component signals, important to study a joint time-frequency (TF) representation, which shows how the frequency content of the signal varies with time. This is done in the field of time-frequency analysis, which is the topic of this thesis.There exist many different joint TF representations for any given signal and choosing an appropriate representation is most often not straight forward. Unfortunately there exist no optimal TF representation for all signals and finding good representations, especially for multi-component signals is a complex problem.In this thesis, methods for obtaining good TF representations, for two types of non-stationary and multi-component signals, and for extracting meaningful information from these representations, are developed. The two types of signals are long, frequency modulated signals and short, transient signals. Even though the types of signals are very different and require very different TF representations, the aim is to resolve components that are close in time, frequency or both. This requires TF representations with high resolution.For the long, frequency modulated signals, a signal adaptive method, which enables automatic comparison between different TF representations, is proposed. For the short, transient signals, a method which finds the TF centres of transient pulses and counts the number of pulses in a signal is presented. An approach for determining the (time) shape of transient pulses is also given.
|
|
| 6. |
- Reinhold, Isabella, et al.
(författare)
-
Non-parametric Envelope Estimation for the Matched Window Reassignment
- 2021
-
Ingår i: 29th European Signal Processing Conference, EUSIPCO 2021 - Proceedings. - 2219-5491. - 9789082797060 ; 2021-August, s. 1920-1924
-
Konferensbidrag (refereegranskat)abstract
- The reassigned spectrogram is a powerful tool for analysing non-stationary signals, and in an ideal setting it gives perfect time and frequency localisation. A method very well suited for oscillating transient signals is the matched window reassignment, which requires a matching window, i.e. the envelope of the transient, to be known or estimated beforehand. This paper proposes a novel method for estimating the envelope of noisy transients, using a non-parametric and computationally efficient approach. The estimated envelopes are used to calculate the matched window reassignment, obtaining estimates of the time-frequency centre of the transients. The reassignment using the estimated envelopes is shown to give good estimates of the time-frequency centres, and good localisation in time and frequency. The novel envelope estimation approach is illustrated on measured marine biosonar data.
|
|
| 7. |
- Reinhold, Isabella, et al.
(författare)
-
Objective detection and time-frequency localization of components within transient signals
- 2018
-
Ingår i: Journal of the Acoustical Society of America. - : Acoustical Society of America (ASA). - 0001-4966. ; 143:4, s. 2368-2378
-
Tidskriftsartikel (refereegranskat)abstract
- An automatic component detection method for overlapping transient pulses in multi-component signals is presented and evaluated. The recently proposed scaled reassignment technique is shown to have the best achievable resolution for closely located Gaussian shaped transient pulses, even in heavy disruptive noise. As a result, the method automatically detects and counts the number of transients, giving the center times and center frequencies of all components with considerable accuracy. The presented method shows great potential for applications in several acoustic research fields, where coinciding Gaussian shaped transients are analyzed. The performance is tested on measured data from a laboratory pulse-echo setup and from a dolphin echolocation signal measured simultaneously at two different locations in the echolocation beam. Since the method requires little user input, it should be easily employed in a variety of research projects.
|
|
| 8. |
- Reinhold, Isabella, et al.
(författare)
-
Optimal time–frequency distributions using a novel signal adaptive method for automatic component detection
- 2017
-
Ingår i: Signal Processing. - : Elsevier BV. - 0165-1684. ; 133, s. 250-259
-
Tidskriftsartikel (refereegranskat)abstract
- Finding objective methods for assessing the performance of time–frequency distributions (TFD) of measured multi-component signals is not trivial. An optimal TFD should have well resolved signal components (auto-terms) and well suppressed cross-terms. This paper presents a novel signal adaptive method, which is shown to have better performance than the existing method, of automatically detecting the signal components for TFD time instants of two-component signals. The method can be used together with a performance measure to receive automatic and objective performance measures for different TFDs, which allows for an optimal TFD to be obtained. The new method is especially useful for signals including auto-terms of unequal amplitudes and non-linear frequency modulation. The method is evaluated and compared to the existing method, for finding the optimal parameters of the modified B-distribution. The performance is also shown for an example set of Heart Rate Variability (HRV) signals.
|
|
| 9. |
- Reinhold, Isabella, et al.
(författare)
-
Quality of Academic Writing for Engineering Students at Lund University
- 2017
-
Ingår i: Transforming patterns through the scholarship of teaching and learning : the 2nd European conference for the scholarship of teaching and learning | Lund, Sweden, June 8-9 2017 - the 2nd European conference for the scholarship of teaching and learning | Lund, Sweden, June 8-9 2017.
-
Konferensbidrag (refereegranskat)abstract
- The development of good writing skills is fundamental not only to publish scientific results, but also to have a deeper understanding of complex subjects. However reports from the Swedish Higher Education Authority from 2012-2015 suggest that engineering students at Lund University have unsatisfactory academic writing skills. Our paper investigates how students of the Faculty of Engineering at Lund University perceive academic writing and the effects of a writing intervention.To obtain information on students’ education in, attitude, perception of its difficulty and their own skill towards academic writing, we constructed a survey which was sent out to engineering students at Lund University. This yielded 62 responses showing that students want more opportunities to develop their writing, but that they grade their writing skill high and the difficulty low.The effects of a simple writing intervention, briefly discussing the importance of good writing practices during a laboratory session, was investigated for one student group. This was done using a random selection of laboratory reports for one control group and one intervention group. The overall quality of the students’ writing was not significantly improved with the intervention. However the quality of the conclusions, the report section with the lowest quality rating, was higher for the intervention group.
|
|
| 10. |
- Reinhold, Isabella
(författare)
-
Spectral Analysis for Signal Detection and Classification : Reducing Variance and Extracting Features
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Spectral analysis encompasses several powerful signal processing methods. The papers in this thesis present methods for finding good spectral representations, and methods both for stationary and non-stationary signals are considered. Stationary methods can be used for real-time evaluation, analysing shorter segments of an incoming signal, while non-stationary methods can be used to analyse the instantaneous frequencies of fully recorded signals. All the presented methods aim to produce spectral representations that have high resolution and are easy to interpret. Such representations allow for detection of individual signal components in multi-component signals, as well as separation of close signal components. This makes feature extraction in the spectral representation possible, relevant features include the frequency or instantaneous frequency of components, the number of components in the signal, and the time duration of the components. Two methods that extract some of these features automatically for two types of signals are presented in this thesis. One adapted to signals with two longer duration frequency modulated components that detects the instantaneous frequencies and cross-terms in the Wigner-Ville distribution, the other for signals with an unknown number of short duration oscillations that detects the instantaneous frequencies in a reassigned spectrogram. This thesis also presents two multitaper methods that reduce the influence of noise on the spectral representations. One is designed for stationary signals and the other for non-stationary signals with multiple short duration oscillations. Applications for the methods presented in this thesis include several within medicine, e.g. diagnosis from analysis of heart rate variability, improved ultrasound resolution, and interpretation of brain activity from the electroencephalogram.
|
|
