Search

In this paper, we propose three-dimensional(3D) visualization of objects in heavy scattering media by using peplography and wavelet transform. Conventional haze removal techniques can remove the light haze in the image by using various image processing algorithms or machine learning techniques. However, they may not provide a clear image under heavy scattering media. On the other hand, peplography can visualize the object by detecting ballistic object photons from heavy scattering media. Then, 3D image can be generated by integral imaging. However, it may not visualize 3D object information accurately because of the noise photons from scattering media. Therefore, the image quality for 3D object visualization may be degraded. To solve this problem, we use the discrete wavelet transform in peplography. It can detect the object photon signals from the scattering media and enhance 3D image contrast ratio by using a specific coefficient threshold technique. To prove our method, we carry out optical experiments and compare results with the conventional haze removal method and peplography by using various image quality metrics such as correlation, structural similarity, and peak signal-to-noise ratio.

Image encryption technique has been actively researched and improved. Double Random Phase Encryption (DRPE), known as the typical way of optical encryption, can easily encrypt images. In the decryption process of DRPE, the complex conjugate of random phase mask used as the encryption key in the Fourier domain is used as the decryption key. As a result, computational DRPE has a vulnerability in image transmission when the random phase mask is stolen. Our proposed method utilizes the random function defined by the sender, and it generates additional random variable to multiply the random phase mask. Finally, our proposed method can enhance the security of the encryption process. However, the random phase mask and random variable information generated by random function are needed to decrypt the image. To send both information, high cost is required in the transmission process. To overcome this problem, our proposed method sends the conjugate random phase mask and random function to the receiver for the decryption process. Finally, our proposed method can enhance security with a low transmission cost., 13th International Conference on ITC Convergence (ICTC2021), October 20-22, 2021, Jeju, Korea

In this paper, we propose an augmented reality head-up display (ARHUD) algorithm and system that can enhance the information visualization for drivers. Recently, Head-up Display (HUD) getting popular in the automotive industry. It can serve the car and other information to the driver efficiently. Therefore, the drivers do not have to move the perspectives to the car dashboard. Finally, it can reduce the accident rate effectively. On the other hand, HUD serves only 2D information, and it is challenging to provide suitable 3D information according to road conditions. To overcome this problem, ARHUD has been developed. ARHUD can generate 3D information and can serve adaptive information for drivers by using a car camera. However, the conventional ARHUD defines content color without considering the road condition. It can cause a critical visualization problem. For example, if the information color and the car color are the same in an actual traffic situation, it is not easy to visualize the content information. In addition, to recognize the road situation in real-time, an image processing machine is needed. However, the conventional machine consumes much electricity, and it can cause vehicle discharge. To solve these problems, we propose an effective processing algorithm and system which can define the content color and recognize the object with high speed without consumes high electricity., 13th International Conference on ITC Convergence (ICTC2021), October 20-22, 2021, Jeju, Korea

We propose a novel monocular depth estimator, which improves the prediction accuracy on human regions by utilizing pose information. The proposed algorithm consists of two networks — PoseNet and DepthNet — to estimate keypoint heatmaps and a depth map, respectively. We incorporate the pose information from PoseNet to improve the depth estimation performance of DepthNet. Specifically, we develop the feature blending block, which fuses the features from PoseNet and DepthNet and feeds them into the next layer of DepthNet, to make the networks learn to predict the depths of human regions more accurately. Furthermore, we develop a novel joint training scheme using partially labeled datasets, which balances multiple loss functions effectively by adjusting weights. Experimental results demonstrate that the proposed algorithm can improve depth estimation performance significantly, especially around human regions. For example, the proposed algorithm improves the depth estimation performance on the human regions of ResNet-50 by 2.8% and 7.0% in terms of $\delta _{1}$ and RMSE, respectively, on the proposed HD + P dataset.

In this study, we demonstrated low leakage current and high mobility thin body (In)GaSb p-FETs. Through the optimization of the V/III ratio during the epitaxial growth, we achieved a highly insulating bottom Al0.95Ga0.05Sb barrier, which eliminates the junction leakage. We also suppressed the interface trap-assisted surface leakage current by introducing In0.53Ga0.47As surface passivation on the GaSb channel. Furthermore, GaSb/InGaSb/GaSb quantum well (QW) channel structure provided significant improvement in effective mobility ( ${\mu }_{\mathrm{ eff}}$ ) characteristics. As a result, the fabricated devices showed the lowest off-leakage current ( ${I} _{\mathrm{ off}}$ ), subthreshold slope ( ${S}$ . ${S}$ .) and high $\mu _{\mathrm{ eff}}$ among reported GaSb p-MOSFETs.

High dynamic range (HDR) image and video technology can provide significant picture quality improvement over the standard dynamic range (SDR). However, when HDR content is represented on an SDR display, dynamic range compression may result in image quality deterioration. To address this problem, we propose an optimized human visual system (HVS) response model-based tone-mapping algorithm to preserve the perceptual responses between the HDR image and its tone-mapped image. First, we measure the HVS response differences using a 2D histogram when an HDR image is displayed on an HDR device and when its tone-mapped image is displayed on an SDR device. Then, we formulate an optimization problem to minimize the differences. By efficiently solving the optimization problem, we obtain an optimal tone-mapping curve. Experimental results on actual displays demonstrate that the proposed algorithm provides superior image quality compared with conventional algorithms in terms of both subjective and objective evaluations.

In this paper, we propose three-dimensional (3D) photon counting double random phase encryption (DRPE) with enhanced visual quality and security level. Conventional 3D photon counting DRPE can quickly encrypt the data by using the 4f optical system and random phase masks with enhanced security because the 3D photon counting technique is used. 3D photon counting DRPE extracts photons from the encrypted data by using statistical methods such as the Poisson random process, and the visual quality of the decrypted data can be enhanced through the 3D reconstruction process. However, it still has a problem to visualize the data when we extract extremely a few photons. To improve the security and the visual quality of the decrypted data, we propose the random amplitude reconstruction process in the encryption stage. Our proposed method reconstructs the amplitude of the encrypted data at a random depth. Thus, the shifting pixel value and depth information can be another important key for decryption through the random process. Therefore, it can effectively decrypt data securely, and the visual quality of the decrypted data can be enhanced. Finally, through the random reconstruction process in the decryption stage, our proposed method can simultaneously enhance the security and visual quality. To verify our proposed method, we carry out the simulation and optical experiment.

This paper reviews the challenge on constrained high dynamic range (HDR) imaging that was part of the New Trends in Image Restoration and Enhancement (NTIRE) workshop, held in conjunction with CVPR 2022. This manuscript focuses on the competition set-up, datasets, the proposed methods and their results. The challenge aims at estimating an HDR image from multiple respective low dynamic range (LDR) observations, which might suffer from under- or over-exposed regions and different sources of noise. The challenge is composed of two tracks with an emphasis on fidelity and complexity constraints: In Track 1, participants are asked to optimize objective fidelity scores while imposing a low-complexity constraint (i.e. solutions can not exceed a given number of operations). In Track 2, participants are asked to minimize the complexity of their solutions while imposing a constraint on fidelity scores (i.e. solutions are required to obtain a higher fidelity score than the prescribed baseline). Both tracks use the same data and metrics: Fidelity is measured by means of PSNR with respect to a ground-truth HDR image (computed both directly and with a canonical tonemapping operation), while complexity metrics include the number of Multiply-Accumulate (MAC) operations and runtime (in seconds)., Comment: CVPR Workshops 2022. 15 pages, 21 figures, 2 tables

Cell segmentation is challenging owing to the existence of various experimental configurations, cell shapes that cannot be mathematically defined, and ambiguous cell boundaries. We propose a cell segmentation method using cell region discriminator and multi-cell discriminator trained using heterogeneous machine-learning techniques such as logistic regression, expectation-maximization, and support vector machine (SVM). The cell-region discriminator identifies the regions where cells are found in images obtained from microscopes via a secondary logistic regressor, and its features use statistical information as well as the distribution of neighbor intensities. The SVM-based multi-cell discriminator determines whether multiple cells are present in the region detected by the cell-region discriminator and whether the region should be divided using the expectation-maximization algorithm. We suggest features for the boundary sectional area and the least square error for cell surface fitting to train the multi-cell discriminator. Using the features and the SVM, the multi-cell discriminator can be trained without overfitting, even for small training data. During this process, the proposed convex cell surface enhances the clustering performance. In experiments, our method based on two discriminators stably divided connected cells even when the contrast between a cell and the background area was small, and it outperformed state-of-the-art methods in terms of cell detection and segmentation accuracy.

In the era of the fourth industrial revolution (Industry 4.0) and the Internet of Things (IoT), real-time data is enormously collected and analyzed from mechanical equipment. By classifying and characterizing the measured signals, the fault condition of mechanical components could be identified. However, most current health monitoring techniques utilize time-consuming and labor-intensive feature engineering, i.e., feature extraction and selection, that are carried out by experts. This paper, on the contrary, deals with an automatic diagnosis method of machine monitoring using a convolutional neural network (CNN) with class activation maps (CAM). A class activation map enables us to discriminate the fault region in the images, thus allowing us to localize the fault precisely. The goal of the paper is to demonstrate how CNN and CAM could be employed to real-world vibration video to characterize the machine’s status, representing normal or fault conditions. The performance of the proposed model is validated with a base-excited cantilever beam dataset and a water pump dataset. This paper presents a novel industrial application by developing a promising method for automatic machine condition-based monitoring.

Knee osteoarthritis (KOA) is a leading cause of disability among elderly adults, and it causes pain and discomfort and limits the functional independence of such adults. The aim of this study was the development of an automated classification model for KOA, based on the Kellgren-Lawrence (KL) grading system, using radiographic imaging and gait analysis data. Gait features highly associated with the radiological severity of KOA identified from our previous study, in addition to radiographic image features extracted from a deep learning network, namely, Inception-ResNet-v2, were exploited using a support vector machine for KOA multi-classification. The area under the curve (AUC) of the receiver operating characteristic curve from KL Grades 0-4 were 0.93, 0.82, 0.83, 0.88, and 0.97, respectively. The sensitivity, precision, and F1-score of the model were 0.70, 0.76, and 0.71, respectively. The proposed model outperformed a common deep learning approach that is based on using only radiographic images as the input data. This result indicates that gait data and radiographic images are complementary with respect to KOA classification, and the use of both data can improve the accuracy of the automated diagnosis of multiclass KOA.

In this paper, a tri-band Gysel power divider (PD) based on zero-degree composite right-/left-handed (CRLH) transmission lines (TLs) is presented. Compared to the generalized design methods in which multiple passbands are realized by replacing conventional impedance transformers with quarter-wavelength CRLH-TLs, the employed zero-degree CRLH-TLs include high design flexibility for the characteristic impedance and phase responses of the composed left-handed and right-handed TLs, which can be used to separately control the bandwidths of the three passbands. In addition, the microstrip-to-slotline transition is utilized to replace the conventional 180° TL, which can expand the realizable range of the three center frequencies and reduce the circuit size, as well. The proposed theory is validated using two designs: One involves a broadband PD with a center frequency of 2.5 GHz, while the other is a tri-band PD centered at 3.5 GHz, 2.5 GHz, and 1.5 GHz. Good agreement between the simulated and measured results are obtained; for the first PD, a fractional 3-dB bandwidth of up to 116% is achieved, whereas, for the tri-band PD, a size reduction of 80% is achieved with good impedance matching and isolation performance.

We present a novel online end-to-end neural diarization system, BW-EDA-EEND, that processes data incrementally for a variable number of speakers. The system is based on the Encoder-Decoder-Attractor (EDA) architecture of Horiguchi et al., but utilizes the incremental Transformer encoder, attending only to its left contexts and using block-level recurrence in the hidden states to carry information from block to block, making the algorithm complexity linear in time. We propose two variants: For unlimited-latency BW-EDA-EEND, which processes inputs in linear time, we show only moderate degradation for up to two speakers using a context size of 10 seconds compared to offline EDA-EEND. With more than two speakers, the accuracy gap between online and offline grows, but the algorithm still outperforms a baseline offline clustering diarization system for one to four speakers with unlimited context size, and shows comparable accuracy with context size of 10 seconds. For limited-latency BW-EDA-EEND, which produces diarization outputs block-by-block as audio arrives, we show accuracy comparable to the offline clustering-based system.

In this paper, we present building entry loss (BEL) measurements on a line-of-sight (LoS) and non-line-of-sight (nLoS) path at 3, 6, and 10 GHz. To investigate impacts of reflected energy (RE) from surrounding environments on the BEL and whole propagation loss, measured and analyzed results are presented according to the presence or absence of the large clutter (a training ship) behind the building. A new parameter RE is proposed for the path loss prediction in the nLoS path and its value is extracted using the measurement results. As a result of the analysis, it was proved that accurate path loss prediction is possible by adding an excess loss (EL) to the sum of clutter loss (CL) and BEL. Therefore, the proposed RE should be considered as a parameter to predict the received power indoors on the nLoS path. Finally, RE measurement and calculation methods are proposed.

In a smart grid, various types of queries such as ad-hoc queries and analytic queries are requested for data. There is a limit to query evaluation based on a single node database engines because queries are requested for a large scale of data in the smart grid. In this paper, to improve the performance of retrieving a large scale of data in the smart grid environment, we propose a DQN-based join order optimization model on Spark SQL. The model learns the actual processing time of queries that are evaluated on Spark SQL, not the estimated costs. By learning the optimal join orders from previous experiences, we optimize the join orders with similar performance to Spark SQL without collecting and computing the statistics of an input data set.