29 November 2023 in Archives of Acoustics
In this article, the authors present the geometry and measurements of the properties of an acoustic metamaterial with a structure composed of multiple concentric rings. CAD models of the structure were developed and subsequently used in numerical studies, which included the study of resonant frequencies using the Lanczos method and an analysis of sound pressure level distribution under plane wave excitation using the finite element method. Subsequently, experimental tests were carried out on models with the same geometry produced with three different materials (PLA, PET-G, and FLEX) using a fused deposition modeling 3D printing technique. These tests included: determining insertion loss for a single model based on tests using the measurement window of a reverberation chamber and determining transmission loss through tests in a semi-anechoic chamber. Sound wave resonance was obtained for frequencies ranging from 1700 to 6000 Hz. Notably, the experimental studies were carried out for the same structure for which numerical tests were conducted. The physical models of a metamaterial were manufactured using three different readily available 3D printing materials. The results of laboratory tests confirm that the created acoustic metamaterial consisting of multi-ring structures reduces noise in medium and high frequencies.
10 May 2023 in Medycyna Pracy
Background: Cognitive efficiency is crucial for many areas of human activity. It affects the employees’ efficiency and safety at the workplace. An important element of the prevention of accidents at work is the appropriate professional selection, which consists in checking the characteristics...
11 July 2025 in International Journal of Electronics and Telecommunications
The following paper presents research on the Automatic Speech Recognition (ASR) methods for the construction of a system to automatically transcribe the medical interview in Polish language during a visit in the clinic. Performance of four ASR models based on Deep Neural Networks (DNN) was evaluated. The applied structures included XLSR-53 large, Quartznet15x5, FastConformer Hybrid Transducer-CTC and Whisper large. The study was conducted on a self-developed speech dataset. Models were evaluated using Word Error Rate (WER), Character Error Rate (CER), Match Error Rate (MER), Word Accuracy (WAcc), Word Information Preserved (WIP), Word Information Lost (WIL), Levenshtein distance, Jaro - Winkler similarity and Jaccard index. The results show that the Whisper model outperformed other tested solutions in the vast majority of the conducted tests. Whisper achieved a WER = 20.84%, where XLSR-53 WER = 67.96%, Quartznet15x5 WER = 76.25%, FastConformer WER = 46.30%. These results show that Whisper needs further adaptation for medical conversations, as current volume of transcription errors is not practically acceptable (too many mistakes in the description of the patient's health description).
5 April 2017 in PRZEGLĄD MECHANICZNY
30 June 2023 in Niepełnosprawność i Rehabilitacja
Celem artykułu było przedstawienie opinii osób z niepełnosprawnością narządu wzroku na temat wyzwań i przeszkód, jakie napotykają podczas codziennego przemieszczania się w przestrzeni miejskiej. Opinie te zostały zebrane podczas bezpośredniego wywiadu przeprowadzonego w grupie 32 osób w wieku 20–40 lat. Grupa ta została równolicznie podzielona ze względu na płeć oraz rodzaj niepełnosprawności narządu wzroku (osoby niewidome vs. osoby słabowidzące). Zorganizowano spotkania w grupach – 4-osobowych, podczas których na podstawie opracowanego scenariusza zadawano pytania odnośnie do poruszania się w przestrzeni miejskiej. Największym problem przy poruszaniu się w przestrzeni miejskiej stanowi lokalizowanie przejścia dla pieszych, barierek i słupków drogowych, hulajnóg, a także trudności z interpretacją numeru autobusu i przystanku komunikacji miejskiej. Dotyczy to zarówno osób niewidomych, jak i słabowidzących.
1 August 2021 in INTER-NOISE and NOISE-CON Congress and Conference Proceedings
Among the factors related to the operation of wind farms, wind turbine noise has to be seen as a source of annoyance for both people living and working near wind farms. A method and a test bench to conduct noise annoyance tests of different types of wind turbine noise in laboratory conditions have been developed. The test bench is based on a multi-channel sound reproducing system using the DANTE network (in which digital acoustic signals are broadcast over Ethernet) and is compiled in the acoustic test chamber. The test bench consists of 19 speakers, including 17 Avantone MixCube studio monitors and 2 LS600 woofers. During the tests a study subject is assessed in terms of efficiency and performance using a computer-based ALS test from the Vienna Test System. The paper describes the test method, the test bench and the results of the pilot studies carried out to assess the impact of wind turbine noise on human performance.