Kwang Myung Jeon

Over the last decade, highly pathogenic avian influenza (HPAI) has severely affected poultry production systems across the globe. In particular, massive pre-emptive depopulation of all poultry within a certain distance has raised concerns regarding animal welfare and food security. Thus, alternative approaches to reducing unnecessary depopulation, such as risk-based depopulation, are highly demanded. This paper proposes a data-driven method to generate a rule table and risk score for each farm to identify preventive measures against HPAI. To evaluate the proposed method, 105 cases of HPAI occurring in a total of 381 farms in Jeollanam-do from 2014 to 2023 were evaluated. The accuracy of preventive measure identification was assessed for each case using both the conventional culling method and the proposed data-driven method. The evaluation showed that the proposed method achieved an accuracy of 84.19%, significantly surpassing the previous 10.37%. The result was attributed to the proposed method reducing the false-positive rate by 83.61% compared with the conventional method, thereby enhancing the reliability of identification. The proposed method is expected to be utilized in selecting farms for monitoring and management of HPAI.

Over the last decade, highly pathogenic avian influenza (HPAI) has severely affected poultry production systems across the globe. In particular, massive pre-emptive depopulation of all poultry within the certain distance has raised concerns regarding animal welfare and food security. Thus, alternative approaches to reducing unnecessary depopulation, such as risk-based depopulation, are highly demanded. This paper proposes a data-driven method to generate a rule table and risk score for each farm to identify preventive measures against HPAI. To evaluate the proposed method, 105 cases of HPAI occurring in a total of 381 farms in Jeollanam-do from 2014 to 2023 were evaluated. The accuracy of preventive measure identification was assessed for each case using both the conventional culling method and the proposed data-driven method. The evaluation showed that the proposed method achieved an accuracy of 84.19%, significantly surpassing the previous 10.37%. The result was attributed to the proposed method reducing the false-positive rate by 83.61% compared with the conventional method, thereby enhancing the reliability of identification. The proposed method is expected to be utilized in selecting farms for monitoring and management of HPAI.

In this paper, a novel speech enhancement method based on a hybrid machine-learning architecture consisting of U-Net and nonnegative matrix factorization (NMF) is proposed. The proposed method attempts to take advantage of both the accurate separation for known noise environments by U-Net and the adaptation to unseen noises by an NMF with an online dictionary learning technique. To merge the two different architectures, a modified U-Net with a temporal activation layer (TAU-Net) is jointly optimized with NMF models that represent universal speech and noise. The proposed method first estimates the temporal activations from the encoder of the proposed TAU-Net. Then, an NMF with online dictionary learning adjusts the initially given temporal activations to suppress their cross-activations due to unseen noises that are unknown in the training phase of TAU-Net. Finally, clean speech is obtained by adjusting temporal activations to the TAU-Net decoder. The effectiveness of the proposed TAU-Net-based speech enhancement method is evaluated in various unseen noise environments. Consequently, the proposed method achieves a substantial improvement with average signal-to-distortion ratios of 2.32 dB and 5.68 dB, which are higher than those of the baseline methods such asspeech enhancement generative adversarial network (SEGAN) and U-Net, respectively.

Deep neural networks (DNNs) have achieved significant advancements in speech processing, and numerous types of DNN architectures have been proposed in the field of sound localization. When a DNN model is deployed for sound localization, a fixed input size is required. This is generally determined by the number of microphones, the fast Fourier transform size, and the frame size. if the numbers or configurations of the microphones change, the DNN model should be retrained because the size of the input features changes. in this paper, we propose a configuration-invariant sound localization technique using the azimuth-frequency representation and convolutional neural networks (CNNs). the proposed CNN model receives the azimuth-frequency representation instead of time-frequency features as the input features. the proposed model was evaluated in different environments from the microphone configuration in which it was originally trained. for evaluation, single sound source is simulated using the image method. Through the evaluations, it was confirmed that the localization performance was superior to the conventional steered response power phase transform (SRP-PHAT) and multiple signal classification (MUSIC) methods.

This paper proposes a sound event detection (SED) method in tunnels to prevent further uncontrollable accidents. Tunnel accidents are accompanied by crashes and tire skids, which usually produce abnormal sounds. Since the tunnel environment always has a severe level of noise, the detection accuracy can be greatly reduced in the existing methods. To deal with the noise issue in the tunnel environment, the proposed method involves the preprocessing of tunnel acoustic signals and a classifier for detecting acoustic events in tunnels. For preprocessing, a non-negative tensor factorization (NTF) technique is used to separate the acoustic event signal from the noisy signal in the tunnel. In particular, the NTF technique developed in this paper consists of source separation and online noise learning. In other words, the noise basis is adapted by an online noise learning technique for enhancement in adverse noise conditions. Next, a convolutional recurrent neural network (CRNN) is extended to accommodate the contributions of the separated event signal and noise to the event detection; thus, the proposed CRNN is composed of event convolution layers and noise convolution layers in parallel followed by recurrent layers and the output layer. Here, a set of mel-filterbank feature parameters is used as the input features. Evaluations of the proposed method are conducted on two datasets: a publicly available road audio events dataset and a tunnel audio dataset recorded in a real traffic tunnel for six months. In the first evaluation where the background noise is low, the proposed CRNN-based SED method with online noise learning reduces the relative recognition error rate by 56.25% when compared to the conventional CRNN-based method with noise. In the second evaluation, where the tunnel background noise is more severe than in the first evaluation, the proposed CRNN-based SED method yields superior performance when compared to the conventional methods. In particular, it is shown that among all of the compared methods, the proposed method with the online noise learning provides the best recognition rate of 91.07% and reduces the recognition error rates by 47.40% and 28.56% when compared to the Gaussian mixture model (GMM)–hidden Markov model (HMM)-based and conventional CRNN-based SED methods, respectively. The computational complexity measurements also show that the proposed CRNN-based SED method requires a processing time of 599 ms for both the NTF-based source separation with online noise learning and CRNN classification when the tunnel noisy signal is one second long, which implies that the proposed method detects events in real-time.

In augmented reality (AR), audio markers can be alternatives to image markers for rendering virtual objects when an AR device camera fails to identify the image marker due to lighting conditions and/or the distance between the marker and device. However, conventional audio markers simply broadcast a rendering queue to anonymous devices, making it difficult to provide specific virtual objects of interest to the user. To overcome this limitation without relying on camera-based sensing, we propose a user-aware audio marker system using low frequency ultrasonic signal processing. The proposed system detects users who stay within the marker using ultrasonic-based object detection, and then it uses ultrasonic communication based on windowed differential phase shift keying modulation in order to send a rendering queue only to those users near the marker. Since the proposed system uses commercial microphones and speakers, conventional telecommunication systems can be employed to deliver the audio markers. The performance of the proposed audio marker system is evaluated in terms of object detection accuracy and communication robustness. First, the object detection accuracy of the proposed system is compared with that of a pyroelectric infrared (PIR) sensor-based system in indoor environments, and it is shown that the proposed system achieves a lower equal error rate than the PIR sensor-based system. Next, the successful transmission rate of the proposed system is measured for various distances and azimuths under noisy conditions, and it is also shown that the proposed audio marker system can successfully operate up to approximately 4 m without any transmission errors, even with 70 dB_SPL ambient noise.

In this paper, we propose a new coordinate-based di-rection-of-arrival (DOA) estimation method, which is applied to several distributed security cameras that have a single microphone each. The proposed method tries to specify the location coordinate of each sound source at regular intervals. To this end, the time differences between multi-channel microphones are first estimated by applying the generalized cross-correlation with phase transform (GCC-PHAT) algorithm to the signals recorded in each pair of microphones. Then, a support vector machine (SVM) is used to classify whether each sound source is located at a closed area constructed by the microphone array. Next, the location of the sound source is estimated as the point at which there is a minimum error between the estimated time difference from the recorded signal to each pair of microphones and the time difference at the specified coordinates using GCC-PHAT. A performance evaluation shows that the classification accuracy of the SVM is 98% and the average distance error is within 10cm.