Signal Enhancement for the SPID Corpus
 |
||
| Contents: | ||
In order to cancel the noise introduced by the ambiance simulation, an adaptive filter based noise cancellation algorithm was applied to the recordings of the SPID corpus. Furthermore, a frequency-domain noise and cross-talk suppression was applied. The two time and time-frequency pots below show examples. The enhanced as well as the original signals are available for the entire corpus parts where the ambiance simualtion was used.
Original Recording
Enhanced Recording
In March 2026, the DSS Chair attended the annual DAGA conference in Dresden. Thanks to the support of the GaS-Club, the student Aylin Kösker was given the opportunity to accompany the chair and participate in the conference from March 23rd to March 26th. As part of the daily poster sessions, she presented the results of her bachelor’s thesis “Machine Learning for the Analysis of Hydrographic Data to Assess the Waterside Accessibility of Port Waters” in the field of Underwater Acoustics. The thesis forms an important basis for an ongoing university research project on the acoustic analysis of sediment properties in harbor areas. The poster session enabled valuable discussions with researchers and conference participants from related research fields.