Journal: Sensors (Basel, Switzerland)

Article Title: Distributed IMU Sensors for In-Field Dynamic Measurements on an Alpine Ski

doi: 10.3390/s24061805

Figure Lengend Snippet: Data flow diagram of the complementary, indirect Kalman filter used for attitude estimation from IMU data. ( A ): Error measurement information is generated through gravity vector estimation from both accelerometer and gyroscope data, hence the complementary filter. ( B ): Indirect Extended Kalman filter equations which operate on attitude error estimations. ( C ): Absolute attitude estimation based on error signals from block B. Note that feedback signals from a previous time step are shown with a dashed line.

Article Snippet: A Kalman filter (KF) implementation (Navigation Toolbox, MATLAB 2023a [ ]) is used to estimate orientation from IMU data.

Techniques: Generated, Plasmid Preparation, Blocking Assay

Journal: Sensors (Basel, Switzerland)

Article Title: Distributed IMU Sensors for In-Field Dynamic Measurements on an Alpine Ski

doi: 10.3390/s24061805

Figure Lengend Snippet: Definitions of terms used in the Indirect Complementary Kalman filter to estimate IMU attitude.

Article Snippet: A Kalman filter (KF) implementation (Navigation Toolbox, MATLAB 2023a [ ]) is used to estimate orientation from IMU data.

Techniques:

Journal: The Journal of the Acoustical Society of America

Article Title: Randomizing spectral cues used to resolve front-back reversals in sound-source localization.

doi: 10.1121/10.0020563

Figure Lengend Snippet: FIG. 2. Spectra of the low-frequency [(A), (B), (D), and (F)] and high-frequency [(C), (E), and (G)] filtered noises, with (A)–(C) indicating the spectra when there was no gammatone filtering (0-dB level change), (D) and (E) when the random-level change was 10 dB (10-dB level change), and (F) and (G) when the random-level change was 20 dB (20-dB level change). The spectrum shown in (A) is for a 200-ms filtered noise, and in all other panels [(B)–(G)], the duration was 2500 ms. The spectra are only examples, as the spectra varied randomly due to the random variation in the level of each gammatone filter. In each case, the amplitudes shown in the figure are scaled to that of the maximum amplitude for the particular noise sample.

Article Snippet: In all other cases, the spectrum of each noise was divided into a series of successive, non-overlapping 1-Cam [equivalent rectangular bandwidth (ERB); see Moore and Glasberg (1983) and Hartmann (1997)] wide filter bands (based on implementation of a gammatone filter bank in MATLAB’s Audio Toolbox).

Techniques: