• 제목/요약/키워드: Spectrogram

검색결과 234건 처리시간 0.026초

CFAR와 합성곱 신경망을 이용한 기두부와 단 분리 시 조각 구분 (Classification of Warhead and Debris using CFAR and Convolutional Neural Networks)

  • 설승환;최인식
    • 한국정보기술학회논문지
    • /
    • 제17권6호
    • /
    • pp.85-94
    • /
    • 2019
  • 기두부와 단 분리 시 조각은 서로 다른 미세 운동을 하므로 스펙트로그램 상에서 미세 도플러 주파수의 형태가 서로 다르게 나타나며 이를 통해 구분이 가능하다. 본 논문에서는 합성곱 신경망(CNN : Convolutional Neural Networks)을 이용하여 기두부와 단 분리 시 조각을 구분하였다. 합성곱 신경망의 입력영상으로는 미세도플러 스펙트로그램을 사용하였다. 또한 기두부와 단 분리 시 조각의 구분성능을 향상시키기 위해 미세 도플러 스펙트로그램에 CA-CFAR(Cell Averaging-Constant False Alarm Rate)를 적용하여 전처리 과정을 수행하였다. 실험 결과, 전처리 과정을 수행하여 획득한 미세 도플러 스펙트로그램을 입력 영상으로 사용하였을 경우, 전처리 과정을 수행하지 않은 미세 도플러 스펙트로그램보다 모든 SNR환경에서 구분 성능이 향상되었다.

비중격 성형술 및 하비잡개 절제술 후 비개존도 측정을 위한 Nasometer와 제1포만트 측정의 유용성 (Significance of Nasometer and First Formant for Nasal Patency After Septoplasty and Turbinoplasty)

  • 진성민;강현국;이경철;박상욱;이성채;이용배
    • 대한후두음성언어의학회지
    • /
    • 제8권2호
    • /
    • pp.161-165
    • /
    • 1997
  • Background : The rhinomanometry and acoustic rhinometry can assess e nasal passage dynamically and statically Recently, analytic methods such as nasometer and sound spectrogram are gaining wide attention to evaluate the nasality objectively. Objectives : firstly to determine if ere was a relationship between the new methods and nasal airway resistance, and secondly to establish if the measurement of nasalance and sound spectrum could be used as an alternative to rhinomanometry and acoustic rhinometry. Materials and Methods : Thirty two patients who underwent either septoplasty and turbinectomy for nasal obstruction were studied. And their ages ranged form 15 to 45 years, with an average of 26.1 years. The rhinomanometry, nasometer, sound spectrogram were performed at preoperative and postoperative 4 weeks day. Results : After operation, subjective symptoms and rhinomanometric results were significantly improved but nasalance and slope of nana, mama and mamma passage had not meningful change. The significnat changes were noted in nasalance and first nasal formant frequency of nasal consonant of velum(angang). Conclusion : Nasometer and sound spectrogram had a limitation for the measure of nasal patency.

  • PDF

Aurally Relevant Analysis by Synthesis - VIPER a New Approach to Sound Design -

  • Daniel, Peter;Pischedda, Patrice
    • 한국소음진동공학회:학술대회논문집
    • /
    • 한국소음진동공학회 2003년도 춘계학술대회논문집
    • /
    • pp.1009-1009
    • /
    • 2003
  • VIPER a new tool for the VIsual PERception of sound quality and for sound design will be presented. Requirement for the visualization of sound quality is a signal analysis modeling the information processing of the ear. The first step of the signal processing implemented in VIPER, calculates an auditory spectrogram by a filter bank adapted to the time- and frequency resolution of the human ear. The second step removes redundant information by extracting time- and frequency contours from the auditory spectrogram in analogy to contours of the visual system. In a third step contours and/or auditory spectrogram can be resynthesised confirming that only aurally relevant information were extracted. The visualization of the contours in VIPER allows intuitively to grasp the important components of a signal. Contributions of parts of a signal to the overall quality can be easily auralized by editing and resynthesising the contours or the underlying auditory spectrogram. Resynthesis of time contours alone allows e.g. to auralize impulsive components separately from the tonal components. Further processing of the contours determines tonal parts in form of tracks. Audible differences between two versions of a sound can be visually inspected in VIPER through the help of auditory distance spectrograms. Applications are shown for the sound design of several interior noises of cars.

  • PDF

능동소나 스펙트로그램 이미지와 CNN을 사용한 표적/비표적 식별 (Target/non-target classification using active sonar spectrogram image and CNN)

  • 김동욱;석종원;배건성
    • 전기전자학회논문지
    • /
    • 제22권4호
    • /
    • pp.1044-1049
    • /
    • 2018
  • CNN(Convolutional Neural Networks)은 동물의 시각정보처리과정을 모델링한 신경망으로 다양한 분야에서 좋은 성능을 보여주고 있다. 본 논문에서는 CNN을 사용하여 능동소나 신호의 스펙트로그램을 분석하고, 표적과 비표적을 식별하는 연구를 수행하였다. 데이터를 표적이 포함된 비율에 따라 8클래스로 구분하고, CNN의 학습에 사용하였다. 신호의 스펙트로그램을 프레임별로 나누어 입력으로 사용한 결과, 표적신호의 위치에서만 표적신호에 해당하는 7개 클래스의 식별 결과가 순차적으로 나타나는 특성을 사용하여 표적과 비표적을 식별해낼 수 있었다.

사전 학습된 딥러닝 모델의 Mel-Spectrogram 기반 기침 탐지를 위한 Attention 기법에 따른 성능 분석 (Attention Modules for Improving Cough Detection Performance based on Mel-Spectrogram)

  • 박창준;김인기;김범준;전영훈;곽정환
    • 한국컴퓨터정보학회:학술대회논문집
    • /
    • 한국컴퓨터정보학회 2023년도 제67차 동계학술대회논문집 31권1호
    • /
    • pp.43-46
    • /
    • 2023
  • 호흡기 관련 전염병의 주된 증상인 기침은 공기 중에 감염된 병원균을 퍼트리며 비감염자가 해당 병원균에 노출된 경우 높은 확률로 해당 전염병에 감염될 위험이 있다. 또한 사람들이 많이 모이는 공공장소 및 실내 공간에서의 기침 탐지 및 조치는 전염병의 대규모 유행을 예방할 수 있는 효율적인 방법이다. 따라서 본 논문에서는 탐지해야 하는 기침 소리 및 일상생활 속 발생할 수 있는 기침과 유사한 배경 소리 들을 Mel-Spectrogram으로 변환한 후 시각화된 특징을 CNN 모델에 학습시켜 기침 탐지를 진행하며, 일반적으로 사용되는 사전 학습된 CNN 모델에 제안된 Attention 모듈의 적용이 기침 탐지 성능 향상에 도움이 됨을 입증하였다.

  • PDF

속도변화에 따른 고속철도차량의 진동 및 소음 특성 (Vibration and noise characteristics of high speed train depending on its speed)

  • 이준석;이시우;고효인;유원희
    • 한국철도학회:학술대회논문집
    • /
    • 한국철도학회 2007년도 춘계학술대회 논문집
    • /
    • pp.73-80
    • /
    • 2007
  • In this paper, the characteristics of noise and vibration of high speed train is analyzed depending on its speed. The speed is a very important parameter because it can affect the interaction between the train and the environment as well as the characteristics of the train itself. To measure its characteristics, we analyzed the signals from microphones and accelerometers which were attached to the passenger car of the high speed train. The signals from each sensor were stored in the recorder, and then analyzed by using the signal processing program. The data from each sensor are analyzed with the spectrogram. From the spectrogram, we found some distinct characteristics of the passenger car. Also, the characteristics of the noise propagation were inferred from the spectrogram.

  • PDF

Neurogram을 이용한 인공와우 자극기법 평가 연구 (Evaluation of Stimulus Strategy for Cochlear Implant Using Neurogram)

  • 양혜진;우지환
    • 대한의용생체공학회:의공학회지
    • /
    • 제34권2호
    • /
    • pp.47-54
    • /
    • 2013
  • Electrical stimulation is delivered to auditory nerve (AN) through the electrodes in cochlear implant system. Neurogram is a spectrogram that includes information of neural response to electrical stimulation. We hypothesized that the similarity between a neurogram and an input-sound spectrogram could show how well a cochlear implant system works. In this study, we evaluated electrical stimulus configuration of CIS strategy using the computational model. The computational model includes stochastic property and anatomical features of cat auditory nerve fiber. To evaluate similarity between a neurogram and an input-sound spectrogram, we calculated Structural Similarity Index (SSIM). The results show that the dynamic range and the stimulation rate per channel influenced SSIM. Finally, we suggested the optimal configuration within the given stimulus CIS. We expect that the results and the evaluating procedure could be employed to improve the performance of a cochlear implant system.

스팩트럼과 스팩트로그램의 이해 (Introduction to the Spectrum and Spectrogram)

  • 진성민
    • 대한후두음성언어의학회지
    • /
    • 제19권2호
    • /
    • pp.101-106
    • /
    • 2008
  • The speech signal has been put into a form suitable for storage and analysis by computer, several different operation can be performed. Filtering, sampling and quantization are the basic operation in digiting a speech signal. The waveform can be displayed, measured and even edited, and spectra can be computed using methods such as the Fast Fourier Transform (FFT), Linear predictive Coding (LPC), Cepstrum and filtering. The digitized signal also can be used to generate spectrograms. The spectrograph provide major advantages to the study of speech. So, author introduces the basic techniques for the acoustic recording, digital signal processing and the principles of spectrum and spectrogram.

  • PDF

A Study on Partial Discharge Diagnostic System for Power Cable using RLCR

  • Park, Keeyoung;Choi, Hyungkee;Lee, Chulhee;Hong, Soomi
    • KEPCO Journal on Electric Power and Energy
    • /
    • 제2권1호
    • /
    • pp.43-47
    • /
    • 2016
  • This system is a diagnosis system that checks whether it causes a partial discharge of a power cable or not. It is to classify normal from abnormal-normal, PD (Partial Discharge) sound through analysis of RLCR (Relative Level Crossing Rate) and spectrogram energy algorithm. Partial discharge diagnostic system has a function that stores PD sound and analyzes the data. The wave shape of PD sound is similar to noise and is systematically generated by partial discharge. Therefore, in this paper, we could discreminate between normal and abnormal case using relative level crossing rate (RLCR) and spectrogram of frequency energy rate.

음소 음향학적 변화 패턴을 이용한 한국어 음성신호의 연속 모음 분할 (Consecutive Vowel Segmentation of Korean Speech Signal using Phonetic-Acoustic Transition Pattern)

  • 박창목;왕지남
    • 한국정보처리학회:학술대회논문집
    • /
    • 한국정보처리학회 2001년도 추계학술발표논문집 (상)
    • /
    • pp.801-804
    • /
    • 2001
  • This article is concerned with automatic segmentation of two adjacent vowels for speech signals. All kinds of transition case of adjacent vowels can be characterized by spectrogram. Firstly the voiced-speech is extracted by the histogram analysis of vowel indicator which consists of wavelet low pass components. Secondly given phonetic transcription and transition pattern spectrogram, the voiced-speech portion which has consecutive vowels automatically segmented by the template matching. The cross-correlation function is adapted as a template matching method and the modified correlation coefficient is calculated for all frames. The largest value on the modified correlation coefficient series indicates the boundary of two consecutive vowel sounds. The experiment is performed for 154 vowel transition sets. The 154 spectrogram templates are gathered from 154 words(PRW Speech DB) and the 161 test words(PBW Speech DB) which are uttered by 5 speakers were tested. The experimental result shows the validity of the method.

  • PDF