Hasil Pencarian  ::  Simpan CSV :: Kembali

"3D sound is a new trend in various media, such as movies, video games, and musicals. Interpolated head-related transfer functions (HRTFs) are a key factor in its production, due to real-time system limitations in storing measured HRTFs. In addition, the interpolation of HRTFs can reduce the need to measure a large amount of HRTFs and the associated effort. In this research, we used the PKU-IOA HRTF Database and covered three interpolation techniques, namely bilinear rectangular, bilinear triangular, and tetrahedral. Bilinear interpolations can be used to compute weights in interpolating measured HRTFs in a linear fashion, with respect to azimuth and elevation angles. Such interpolations have been proposed for three measurement points that form a triangle or for four measurement points that form a rectangle, surrounding the HRTF at a desired point. These geometrical approaches compute weights from a distance of the desired point from each measurement point. Tetrahedral interpolation, meanwhile, is a technique for HRTF measurements in 3D (i.e. azimuth, elevation, and distance) using barycentric weights. Based on our experiments, 3D tetrahedral interpolation results in the best average mean square error (MSE) of 3.72% for minimum phase head related impulse responses (HRIRs) and best average spectral distortion (SD) of 2.79 dB for magnitude HRTFs, compared to 2D bilinear interpolations (i.e. rectangular and triangular interpolation). Regarding the latter, bilinear rectangular interpolation generally performs better than the triangular variety. Additionally, the use of minimum phase HRIRs as input data results in more optimal interpolated data than magnitude HRTFs. We therefore propose an optimal framework for obtaining estimated HRIRs by interpolating minimum phase HRIRs using tetrahedral interpolation. Such HRIRs have been simulated to produce virtual 3D moving sound in a horizontal plane with a difference of 2.5o of azimuth angle. The simulated moving sound that is heard moves naturally in a clockwise direction from an azimuth angle of 0o to 360o."

"3D sound is a new trend in various media, such as movies, video games, and musicals. Interpolated head-related transfer functions (HRTFs) are a key factor in its production, dueto real-time system limitations in storing measured HRTFs. In addition, theinterpolation of HRTFs can reduce the need to measure a large amount of HRTFs andthe associated effort. In this research, we used the PKU-IOA HRTF Database andcovered three interpolation techniques, namely bilinear rectangular, bilineartriangular, and tetrahedral. Bilinear interpolations can be used to computeweights in interpolating measured HRTFs in a linear fashion, with respect toazimuth and elevation angles. Such interpolations have been proposed for threemeasurement points that form a triangle or for four measurement points thatform a rectangle, surrounding the HRTF at a desired point. These geometricalapproaches compute weights from a distance of the desired point from eachmeasurement point. Tetrahedral interpolation, meanwhile, is a technique forHRTF measurements in 3D (i.e. azimuth, elevation, and distance) using barycentric weights. Based on our experiments, 3D tetrahedralinterpolation results in the best average mean square error (MSE) of 3.72% for minimum phase head related impulse responses (HRIRs) and best average spectral distortion (SD) of2.79 dB for magnitude HRTFs, compared to 2D bilinear interpolations (i.e.rectangular and triangular interpolation). Regarding the latter, bilinearrectangular interpolation generally performs better than the triangularvariety. Additionally, the use of minimum phase HRIRs as input data results inmore optimal interpolated data than magnitude HRTFs. We therefore propose anoptimal framework for obtaining estimated HRIRs by interpolating minimum phaseHRIRs using tetrahedral interpolation. Such HRIRs have been simulated toproduce virtual 3D moving sound in a horizontal plane with a difference of 2.5oof azimuth angle. The simulated moving sound that is heard moves naturally in aclockwise direction from an azimuth angle of 0o to 360o."

"As this monograph shows, the purpose of cardinal spline interpolation is to bridge the gap between the linear spline and the cardinal series. The author explains cardinal spline functions, the basic properties of B-splines, including B- splines with equidistant knots and cardinal splines represented in terms of B-splines, and exponential Euler splines, leading to the most important case and central problem of the book-- cardinal spline interpolation, with main results, proofs, and some applications. Other topics discussed include cardinal Hermite interpolation, semi-cardinal interpolation, finite spline interpolation problems, extremum and limit properties, equidistant spline interpolation applied to approximations of Fourier transforms, and the smoothing of histograms."

"The UN Volume (Bogue, Arriaga and Anderton, l993)on population methodology has elaborated different methods forinterpolation of population data. There were two methods ininterpolating empirical data frequently discussed, that are mid-point method and cumulation-differencing method In some casesmid-point method was not recommended and cumulation-differencing method was recommended only on the basis of limiteddata. This study tries to explore other methods in manipulatingpopulation data for different nature of data. The paper when atone instance, finds that the method which was not recommended bythe said volume was acceptable using different nature of data. Inother case, the method which was recommended in the volume was_,round more sound in respect of other kind of data In addition,some new types of data have also been tried and appropriateinterpolation formulae were recommended."

"Penelitian ini melanjutkan dari penelitian sebelumnya mengenai pengenalan wajah tiga dimensi dengan sudut pandang vertikal-horisontal yang beragam. Penelitian ini menggunakan Jaringan Syaraf Tiruan (JST) Hemispheric Structure of Hidden Layer (HSHL). HSHL adalah struktur JST dengan lapis tersembunyi berbentuk hemisfer. Pada penelitian sebelumnya, informasi sudut pandang yang digunakan HSHL dalam proses pengujian diberikan secara manual dan penelitian tersebut memberikan hasil yang bagus.Pada penelitian ini, penulis mencoba untuk membandingkan antara JST HSHL dengan sudut wajah diketahui sebelumnya dengan JST HSHL menggunakan interpolasi spline sebagai pengenal sudut wajah. Selain itu, penulis juga menggunakan Principal Component Analysis untuk ekstraksi ciri dari data masukan.Penelitian ini menyimpulkan bahwa pengenalan wajah menggunakan JST HSHL dengan interpolasi spline sebagai pengenal sudut wajah memberikan tingkat pengenalan yang tidak jauh berbeda dengan JST HSHL yang sudut wajah diketahui. Oleh karena itu, kombinasi antara JST HSHL dengan interpolasi spline memungkinkan untuk diterapkan dalam aplikasi dunia nyata.

---

This research continues from previous researches about 3D face recognition with variation of vertical-horizontal angle face. This research uses Hemispheric Structure of Hidden Layer (HSHL). In the previous research, angle face information is known by the system and the recognition rate is high.In this research, writer try to compare HSHL with known angle face and HSHL with spline interpolation to guess the angle face. Writer also uses Principal Component Analysis to extract features from input data.This research concludes that HSHL with spline interpolation to guess the agle face yields a competitive result with HSHL with known angle face. Therefore, combination of HSHL and spline interpolation is possible to be applied in the real world."

"This book is primarily devoted to topics related to matrix versions of classical interpolation and moment problems. The major themes include Schur analysis of nonnegative Hermitian block Hankel matrices and the construction of Schur-type algorithms. This book also covers a number of recent developments in orthogonal rational matrix functions, matrix-valued Carathéodory functions and maximal weight solutions for particular matricial moment problems on the unit circle.​"

"Aktivitas domestik masyarakat di pesisir dan kepulauan merupakan salah satu faktor yang dapat mempengaruhi kualitas dan kuantitas dari ketersediaan air tanah. Salah satu parameter limbah cair rumah tangga yaitu pencemar deterjen. Tujuan penelitian ini adalah untuk menganalisis sebaran spasial konsentrasi deterjen dan risiko cemaran deterjen pada air minum masyarakat pulau Kodingareng Lompo. Penelitian ini merupakan jenis penelitian kuantitatif dengan pendekatan deskriptif. Metode yang digunakan adalah model sebaran konsentrasi deterjen dengan interpolasi kriging dan pendekatan Analisis Risiko Kesehatan Lingkungan (ARKL). Sebaran kandungan deterjen tertinggi pada air sumur terdapat pada RW 2 dan RW 3 sebesar 2,384-2,98 mg/l, sedangkan sebaran kandungan deterjen pada air sumur terendah juga terdapat pada RW 2 dan RW 3, dan RW 6 sebesar 0,005-0,0088 mg/L. Berdasarkan pemeriksaan yang dilakukan, konsentrasi tertinggi terdapat pada sampel air minum 4 dan 5, yaitu sebesar 2,98 mg/l. Sedangkan sampel air minum 2 memiliki konsentrasi cemaran deterjen paling rendah, yaitu 0,005 mg/l. Hasil perhitungan ARKL yang dilakukan terhadap konsentrasi cemaran deterjen pada air minum menunjukkan tingkat risiko yang tinggi (RQ >1) pada tiga responden dan tergolong tidak aman untuk konsumsi air minum.

---

Domestic activities in coastal and island communities are one of the factors that can affect the quality and quantity of groundwater availability. One of the parameters of household wastewater is detergent pollutants. The purpose of this study was to analyze the risk of detergent contamination in drinking water in the community of Kodingareng Lompo Island. This research is a type of quantitative research with a descriptive approach. The method used are the model of the distribution of detergent concentrations by interpolating kriging and the Environmental Health Risk Analysis (ARKL) approach. The highest distribution of detergent content in well water is in RW 2 and RW 3 of 2,384 - 2,98 mg/l, while the distribution of detergent content in the lowest well water is also found in RW 2 and RW 3, and RW 6 of 0,005 - 0,0088 mg/l. Based on the examination conducted, the highest concentration was found in drinking water samples 4 and 5, which was equal to 2.98 mg/l. while the drinking water sample 2 has the lowest concentration of detergent contamination, which is 0.005 mg/l. The results of ARKL calculations conducted on the concentration of detergent contamination in drinking water showed a high level of risk (RQ> 1) in three respondents and classified as unsafe for drinking water consumption."

"Citra tipikal hitam-putih merupakan citra kontinyu karena gray level shade-nya bcrgabting secara kontinyu dalam intensitas dan spatiality.Dalam suatu pemrosesan citra dilakukan proses kuantisasi, dimana citra terkuantisasi tidak kontinyu, namun tersusun dari piksel diskrit.Resolusi citra didefinisikan sebagai kemampuan untuk memproduksi citra dengan kualitas visual dan dapat dibandingkan dcngan citra kontinyu aslinya. Diperlukan jumlah piksel nicmadai untuk penampilan kontinyu spasial dan juga jumlah gray level untuk penampilan depth of field dan kontras. Piksel merupakan elemen utama dalam resolusi citra diskrit.Dalam tesis ini diusulkan peningkatan resolusi citra untuk pencitraan medik memakai Filter Infinite Impulse Response. Tiga teknik peningkatan resolusi citra yang dikenal, masing-masing : (1) Pendekatan frekuensi/Fast Fourier Transform dari teorema sampling, (2) Pendekatan spasial memakai fungsi interpolasi; dan (3) pendekatan alternatif memakai Infinite Impulse Response Filter (mirip dengan pendekatan FFT, namun lebih mudah dalam pelaksanaan).Peningkatan resolusi citra melalui FFT dilakukan dengan, pertama, memakai sifat spektrum yang akan cepat mcnurun akibat meningkatnya frekuensi dan kedua, memakai teorema sampling untuk obyek oversumpled.Peningkatan resolusi citra memakai interpolasi spasial dilakukan dengan mengkonvolusikan citra yang sudah diekspansi 2 N x 2 N dengan operator interpolasi. Hasilnya tampak agak buram, namun hal ini dapat di-deblur memakai high pass filter.Peningkatan resolusi citra memakai Infinite Impulse Response fitter, meningkatkan ukuran citra melalui replikasi piksel dan garis horisontal. Effeknya, terjadi peningkatan perioda sampling pada arah horisontal maupun vertikal. Pendekatan ini sama dengan pendekatan FFT dan langkah-langkah tersebut di atas sama dengan menerapkan fungsi interpolasi square.Kualitas citra yang dihasilkan sama dengan pada pendekatan FFT.

---

Typical black and white images are called continuous - tone images because the shade of gray level blend continuosly both in intensity (level) and spatiality (area). The quantized images is not continuous, but is composed of discrete pixels with each having a discrete gray level assigned to it. Image resolution may be defined as the ability to reproduce images with a visual quality that is comparable to the continuous-tone original.This requires a suffient number of pixels to give the appearance of spatial continuity and a sufficient number of gray levels to give the appearance of depth and contrast. The pixel is the basic element.of picture resolution in the discrete pixel system.In the next discussion, three means of doubling image resolution are compared respectively ; (1) The Fast Fourier Transform Frequency approach of the sampling theorem, (2) through spatial approach using interpolation functions; and (3) an alternative approach using Infinite Impulse Response filter, which is similiar to the FFT approach, but much simpler to implement.Doubling image resolution through FFT are implemented by using, first, the rapidly decreasing spectrum with increasing frequency and second, the whittakershannon sampling theorem for oversampled objects.Doubling using spatial interpolation are held by convolving the stretched 2 N x 2 N image with an interpolation operator. The result is slightly blurry, it is, however, possible to deblur through high-pass filtering.Doubling image resolution using Infinite Impulse Respons filter, double the image size by replication of pixels and horizontal lines this has the effect of increasing the horizontal and vertical sampling periods to twice their values. The approach is similar to the FFT approach and the step mentioned above is equivalent to applying the square interpolation function. The enlarged image obtained should be similar to the one obtained from the FFT approach."

"Pelacakan orang banyak pada video berdasarkan hasil deteksi orang pada setiap frame merupakan problem yang menantang karena kompleksitas yang dimilikinya. Kesalahan deteksi orang pada setiap frame akan menyebabkan kesalahan pelacakan orang pada keseluruhan video. Pada penelitian ini, diusulkan metode pelacakan yang dapat meminimalkan propagasi kesalahan dari kesalahan deteksi dengan waktu pelacakan yang tidak terlalu lama. Penelitian ini menggunakan deep convolutional neural network (DCNN) seperti Faster-RCNN dan RetinaNet sebagai detektor objek dan algoritma Hungarian sebagai metode asosiasi antar orang-orang yang terdeteksi di setiap frame. Matriks masukan untuk algoritma Hungarian terdiri dari kedekatan vektor ciri DCNN yang dihasilkan oleh Siamese Network, jarak titik tengah bounding box, dan perbandingan irisan-gabungan (IoU) dari bounding box. Pada tahap akhir dilakukan interpolasi terhadap hasil pelacakan. Metode yang diusulkan menghasilkan MOTA 61.0 pada dataset benchmark pelacakan orang banyak MOT16.

---

Multiple object (human) tracking in video based on object detection in every frame is a challenging problem due to its complexity. Error in the detection phase will cause error in the tracking phase. In this research, a multiple human tracking method is proposed to minimize the error propagation. The method uses deep convolutional neural network (DCNN) such as Faster-RCNN and RetinaNet as object detector and Hungarian algorithm as association method among detected humans in consecutive frames. The input matrix for Hungarian algorithm consists of the similarity of DCNN feature vector resulted from Siamese network, the distance of bounding box centers, and bounding box intersection of union (IoU). In the last step, interpolation is applied to the tracking result. The proposed method achieves 61.0 MOTA in multiple object tracking benchmark MOT16."