Background Noise Removal Technique using Deep Learning Segmentation Network without Segmentation Map

Audio data acquired at industrial manufacturing sites often include unexpected background noise. Since the performance of data-driven models can be worse by background noise. Therefore, it is important to get rid of unwanted background noise. There are two main techniques for noise canceling in a traditional manner. One is Active Noise Canceling (ANC), which generates an inverted phase of the sound that we want to remove. The other is Passive Noise Canceling (PNC), which physically blocks the noise. However, these methods require large device size and expensive cost. Thus, we propose a deep learning-based noise canceling method. This technique was developed using audio imaging technique and deep learning segmentation network. However, the proposed model only needs the information on whether the audio contains noise or not. In other words, unlike the general segmentation technique, a pixel-wise ground truth segmentation map is not required for this method. We demonstrate to evaluate the separation using pump sound of MIMII dataset, which is open-source dataset.

Design of noise barriers for the mitigation of construction noise

Noise pollution is a major problem in many major cities in particular a small island state like Singapore with residential buildings very close to the major trunk roads and expressways. The problem is aggravated by the ongoing city redevelopment and construction of new mass rapid transit lines. Construction noise is therefore a common theme of public complaints and therefore there is an increased interest in the development of more effective mitigation measure for construction noise. In this work, a Flat-tip jagged-edge profile was investigated and applied on the edge of a cantilever (slanted up for 45 degrees, facing the noise source) which was mounted at the top of a passive noise barrier. Besides the numerical simulations, the full sized prototypes were also experimentally tested on a construction sites with noise generated by a boring machine. Both numerical simulations and experimental results showed that this barrier with a slanted Flat-tip jagged cantilever would perform better than the traditional barrier having a Straight-edge cantilever of same height, with a maximum additional attenuation of 5.0 dBA experimentally obtained. The barrier with slanted Flat-tip jagged cantilever could also extend the shadow zone behind the barrier to higher levels of the building.

Experimental/Numerical Acoustic Assessment of Aircraft Seat Headrests Based on Electrospun Mats

The work proposes a methodology for the assessment of the performances of Passive Noise Control (PNC) for passenger aircraft headrests with the aim of enhancing acoustic comfort. Two PNC improvements of headrests were designed to reduce the Sound Pressure Level (SPL) at the passengers’ ears in an aircraft cabin during flight; the first was based on the optimization of the headrest shape, whereas the second consisted of partially or fully covering the headrest surface with a new highly sound-absorbing nanofibrous textile. An experimental validation campaign was conducted in a semi-anechoic chamber. A dummy headrest was assembled in different configurations of shape and materials to assess the acoustic performances associated to each set up. In parallel, simulations based on the Boundary Element Method (BEM) were performed for each configuration and an acceptable correlation between experimental and numerical results was obtained. Based on these findings, general guidelines were proposed for the acoustical design of advanced headrests.

Passive noise control in buildings: An engineering case study of ducted systems

The most common noise sources in buildings are related to Heating, Ventilating and Air Conditioning (HVAC) systems, plumbing systems, electrical systems and exterior sources. Passive Noise Control (PNC) techniques in buildings have been implemented in several ways. The aim of this work is to analyses the use of silencer to attenuate the noise in the ducts that are part of the ventilation systems in buildings, internal combustion systems, fans, gas conduction systems, boilers, etc. The main objective of a silencer is to reduce the transmission of noise, disturbing as little as possible the circulation of gas or liquid. In the first instance, the silencers are classified as reactive and dissipative, depending on whether the attenuation of the noise is produced by reflective or dissipative mechanisms, respectively. In a reactive silencer, the losses occur essentially due to the reflections of the sound waves in impedance discontinuities, such as widening or narrowing of the tube. In dissipative silencers, the flow is in contact with a large surface of absorbent material. The attenuation of the noise is then produced by visco-thermal losses in the porous material. In this work, a practical issue will be addressed with a noise reduction of 19 dBA in 60 Hz. Practical application Noise is a current issue in residential areas that could lead to health problems for people. The origin of these noises within buildings is very diverse, one of them is produced by ducts. Appling the PNC technique in modern building construction would be a good prevention practice. For this reason, in this project a PNC system was carried out in the ducts of a residential building, which could be used as a praiseworthy solution, avoiding problems for the inhabitants of these spaces.