Matching Items (194)
Filtering by
- Creators: Yang, Yezhou
- Creators: Ryan, Russell
- Member of: ASU Electronic Theses and Dissertations
Description
The performance of most of the visual computing tasks depends on the quality of the features extracted from the raw data. Insightful feature representation increases the performance of many learning algorithms by exposing the underlying explanatory factors of the output for the unobserved input. A good representation should also handle anomalies in the data such as missing samples and noisy input caused by the undesired, external factors of variation. It should also reduce the data redundancy. Over the years, many feature extraction processes have been invented to produce good representations of raw images and videos.
The feature extraction processes can be categorized into three groups. The first group contains processes that are hand-crafted for a specific task. Hand-engineering features requires the knowledge of domain experts and manual labor. However, the feature extraction process is interpretable and explainable. Next group contains the latent-feature extraction processes. While the original feature lies in a high-dimensional space, the relevant factors for a task often lie on a lower dimensional manifold. The latent-feature extraction employs hidden variables to expose the underlying data properties that cannot be directly measured from the input. Latent features seek a specific structure such as sparsity or low-rank into the derived representation through sophisticated optimization techniques. The last category is that of deep features. These are obtained by passing raw input data with minimal pre-processing through a deep network. Its parameters are computed by iteratively minimizing a task-based loss.
In this dissertation, I present four pieces of work where I create and learn suitable data representations. The first task employs hand-crafted features to perform clinically-relevant retrieval of diabetic retinopathy images. The second task uses latent features to perform content-adaptive image enhancement. The third task ranks a pair of images based on their aestheticism. The goal of the last task is to capture localized image artifacts in small datasets with patch-level labels. For both these tasks, I propose novel deep architectures and show significant improvement over the previous state-of-art approaches. A suitable combination of feature representations augmented with an appropriate learning approach can increase performance for most visual computing tasks.
The feature extraction processes can be categorized into three groups. The first group contains processes that are hand-crafted for a specific task. Hand-engineering features requires the knowledge of domain experts and manual labor. However, the feature extraction process is interpretable and explainable. Next group contains the latent-feature extraction processes. While the original feature lies in a high-dimensional space, the relevant factors for a task often lie on a lower dimensional manifold. The latent-feature extraction employs hidden variables to expose the underlying data properties that cannot be directly measured from the input. Latent features seek a specific structure such as sparsity or low-rank into the derived representation through sophisticated optimization techniques. The last category is that of deep features. These are obtained by passing raw input data with minimal pre-processing through a deep network. Its parameters are computed by iteratively minimizing a task-based loss.
In this dissertation, I present four pieces of work where I create and learn suitable data representations. The first task employs hand-crafted features to perform clinically-relevant retrieval of diabetic retinopathy images. The second task uses latent features to perform content-adaptive image enhancement. The third task ranks a pair of images based on their aestheticism. The goal of the last task is to capture localized image artifacts in small datasets with patch-level labels. For both these tasks, I propose novel deep architectures and show significant improvement over the previous state-of-art approaches. A suitable combination of feature representations augmented with an appropriate learning approach can increase performance for most visual computing tasks.
ContributorsChandakkar, Parag Shridhar (Author) / Li, Baoxin (Thesis advisor) / Yang, Yezhou (Committee member) / Turaga, Pavan (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2017
Description
This research paper focuses on selected recordings of the Etudes of Claude Debussy. It provides a comparative study of these recordings.
There are some dissertations on the topic of Debussy’s Etudes. Most of them are about performance-related aspects such as fingerings, pedaling, or technical guidelines. Some of the dissertations examine compositional analyses, discussing harmony, texture, rhythmic structure, motivic development, etc. There also is a dissertation that makes a comparative study of the etude genre in Chopin and Debussy. Since there is no research yet on the recordings of Debussy’s Etudes, this may be a meaningful contribution to research. Debussy’s Douze Études are technically difficult to play, but the technical problems are always subordinated to musical beauty and variety in this work. This research is concerned with the sound of the music as achieved by a variety of performers.
Nine representative pianists from various schools and traditions are chosen: Michel Béroff, Aldo Ciccolini, Walter Cosand, Walter Gieseking, Werner Haas, Yvonne Loriod, Jean-Yves Thibaudet, Mitsuko Uchida and Yevgeny Yontov. In this project, the focus is on listening to the selected recordings, making comparisons and summarizing certain performance-related aspects of them. Each etude is discussed individually in order to make a comprehensive study of different aspects of the selected recordings. In the last chapter of this paper, conclusions are drawn about the different performance features of the pianists examined according to previous analyses.
This research seeks to encourage performances of Debussy’s Etudes, to aid pianists in obtaining interpretative ideas from the different recordings and finally to benefit their own performances.
There are some dissertations on the topic of Debussy’s Etudes. Most of them are about performance-related aspects such as fingerings, pedaling, or technical guidelines. Some of the dissertations examine compositional analyses, discussing harmony, texture, rhythmic structure, motivic development, etc. There also is a dissertation that makes a comparative study of the etude genre in Chopin and Debussy. Since there is no research yet on the recordings of Debussy’s Etudes, this may be a meaningful contribution to research. Debussy’s Douze Études are technically difficult to play, but the technical problems are always subordinated to musical beauty and variety in this work. This research is concerned with the sound of the music as achieved by a variety of performers.
Nine representative pianists from various schools and traditions are chosen: Michel Béroff, Aldo Ciccolini, Walter Cosand, Walter Gieseking, Werner Haas, Yvonne Loriod, Jean-Yves Thibaudet, Mitsuko Uchida and Yevgeny Yontov. In this project, the focus is on listening to the selected recordings, making comparisons and summarizing certain performance-related aspects of them. Each etude is discussed individually in order to make a comprehensive study of different aspects of the selected recordings. In the last chapter of this paper, conclusions are drawn about the different performance features of the pianists examined according to previous analyses.
This research seeks to encourage performances of Debussy’s Etudes, to aid pianists in obtaining interpretative ideas from the different recordings and finally to benefit their own performances.
ContributorsJiang, Yuan (Author) / Cosand, Walter (Thesis advisor) / Rogers, Rodney (Committee member) / Ryan, Russell (Committee member) / Arizona State University (Publisher)
Created2018
Description
Our world has become smaller due to globalization and frequent cultural exchange between different countries. As a result, classical music is becoming increasingly global. There are a significant number of Chinese composers, including Tan Dun, Chen Yi, Zhou Long, and Bright Sheng, who have gained international attention. For a modern performer, familiarity with music outside of the Western canon is increasingly important.
Bright Sheng is an internationally renowned Chinese-American composer who blends the heritage of traditional Chinese musical elements, traditional instruments, Chinese Opera and folk melodies with Western musical techniques. He infuses Chinese character into his works and introduces Chinese music to the Western classical music world.
In this paper, I discuss two of Bright Sheng’s pieces: A Night at the Chinese Opera and Three Fantasies. Both works were composed in 2005 and are the only two compositions he wrote for violin and piano. Most pianists are not familiar with how to transfer or imitate the sounds of traditional Chinese instruments on Western musical instruments. The paper examines traditional Chinese techniques for Western instruments from A Night in Chinese Opera. Three Fantasies contains three distinct musical characters related to different musical elements from different regions of China. I explore the traditional musical forms from Three Fantasies and offer practical suggestions for performance practice.
This document provides Bright Sheng’s biography, educational background, influences, and compositional style. It also features the inspirations for both pieces, a detailed analysis of both scores including a structural outline, discussion of compositional style, usage of rhythm and timbre and explanation of special techniques. This document also serves as an interpretative guide to each composition, including story outlines, suggestions for practice strategies, aesthetic considerations, rehearsal techniques and performance considerations.
The research for this paper is based on personal interview and coaching with Bright Sheng and analysis from the published scores for A Night at the Chinese Opera and Three Fantasies by G. Schirmer, Inc. I hope that this document will be a comprehensive performers’ guide to both works and serve as an explanation and promotion of Chinese classical music to a larger audience.
Bright Sheng is an internationally renowned Chinese-American composer who blends the heritage of traditional Chinese musical elements, traditional instruments, Chinese Opera and folk melodies with Western musical techniques. He infuses Chinese character into his works and introduces Chinese music to the Western classical music world.
In this paper, I discuss two of Bright Sheng’s pieces: A Night at the Chinese Opera and Three Fantasies. Both works were composed in 2005 and are the only two compositions he wrote for violin and piano. Most pianists are not familiar with how to transfer or imitate the sounds of traditional Chinese instruments on Western musical instruments. The paper examines traditional Chinese techniques for Western instruments from A Night in Chinese Opera. Three Fantasies contains three distinct musical characters related to different musical elements from different regions of China. I explore the traditional musical forms from Three Fantasies and offer practical suggestions for performance practice.
This document provides Bright Sheng’s biography, educational background, influences, and compositional style. It also features the inspirations for both pieces, a detailed analysis of both scores including a structural outline, discussion of compositional style, usage of rhythm and timbre and explanation of special techniques. This document also serves as an interpretative guide to each composition, including story outlines, suggestions for practice strategies, aesthetic considerations, rehearsal techniques and performance considerations.
The research for this paper is based on personal interview and coaching with Bright Sheng and analysis from the published scores for A Night at the Chinese Opera and Three Fantasies by G. Schirmer, Inc. I hope that this document will be a comprehensive performers’ guide to both works and serve as an explanation and promotion of Chinese classical music to a larger audience.
ContributorsJiang, Zhou (Author) / Ryan, Russell (Thesis advisor) / Campbell, Andrew (Committee member) / Carpenter, Ellon (Committee member) / Arizona State University (Publisher)
Created2018
Description
Even in the most despondent situations, the arts find a way to flourish. This research document examines the vocal music that Viktor Ullmann composed in the concentration camp-ghetto of Theresienstadt in German-occupied Czechoslovakia, and the notable aspects of his compositional writing style. Although his opera Der Kaiser von Atlantis has been performed globally, the remainder of his oeuvre has rarely been recorded or performed. Singers often shy away from twentieth-century composers such as Ullmann, with claims that the music is not lyrical or relatable. Perhaps the irregularity of the meter, rhythms, or intervals seem too daunting for many to consider attempting a performance. With Confined, But Not Silenced: Vocal Music of Viktor Ullmann from the Theresienstadt Ghetto, I hope to open the door to music that is both accessible and uniquely beautiful. Not intended as a performance guide, this document aims instead at unearthing the appeal of music that is often perceived as unusual and difficult to perform through analysis that emphasizes relatable aspects of the compositions. By exposing colleagues to relatable music by a composer such as Ullmann, that has not normally been integrated in the canon, the boundaries of standard vocal repertoire can be expanded to include unconventional contemporary compositions. In addition to highlighting specific musical examples, Confined, But Not Silenced focuses on music’s positive effects for World War II prisoners in Theresienstadt.
ContributorsGoglia, Adrienne Ruth (Author) / FitzPatrick, Carole (Thesis advisor) / Feisst, Sabine (Committee member) / Ryan, Russell (Committee member) / Arizona State University (Publisher)
Created2018
Description
Topological methods for data analysis present opportunities for enforcing certain invariances of broad interest in computer vision: including view-point in activity analysis, articulation in shape analysis, and measurement invariance in non-linear dynamical modeling. The increasing success of these methods is attributed to the complementary information that topology provides, as well as availability of tools for computing topological summaries such as persistence diagrams. However, persistence diagrams are multi-sets of points and hence it is not straightforward to fuse them with features used for contemporary machine learning tools like deep-nets. In this paper theoretically well-grounded approaches to develop novel perturbation robust topological representations are presented, with the long-term view of making them amenable to fusion with contemporary learning architectures. The proposed representation lives on a Grassmann manifold and hence can be efficiently used in machine learning pipelines.
The proposed representation.The efficacy of the proposed descriptor was explored on three applications: view-invariant activity analysis, 3D shape analysis, and non-linear dynamical modeling. Favorable results in both high-level recognition performance and improved performance in reduction of time-complexity when compared to other baseline methods are obtained.
The proposed representation.The efficacy of the proposed descriptor was explored on three applications: view-invariant activity analysis, 3D shape analysis, and non-linear dynamical modeling. Favorable results in both high-level recognition performance and improved performance in reduction of time-complexity when compared to other baseline methods are obtained.
ContributorsThopalli, Kowshik (Author) / Turaga, Pavan Kumar (Thesis advisor) / Papandreou-Suppappola, Antonia (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2017
Description
Image Understanding is a long-established discipline in computer vision, which encompasses a body of advanced image processing techniques, that are used to locate (“where”), characterize and recognize (“what”) objects, regions, and their attributes in the image. However, the notion of “understanding” (and the goal of artificial intelligent machines) goes beyond factual recall of the recognized components and includes reasoning and thinking beyond what can be seen (or perceived). Understanding is often evaluated by asking questions of increasing difficulty. Thus, the expected functionalities of an intelligent Image Understanding system can be expressed in terms of the functionalities that are required to answer questions about an image. Answering questions about images require primarily three components: Image Understanding, question (natural language) understanding, and reasoning based on knowledge. Any question, asking beyond what can be directly seen, requires modeling of commonsense (or background/ontological/factual) knowledge and reasoning.
Knowledge and reasoning have seen scarce use in image understanding applications. In this thesis, we demonstrate the utilities of incorporating background knowledge and using explicit reasoning in image understanding applications. We first present a comprehensive survey of the previous work that utilized background knowledge and reasoning in understanding images. This survey outlines the limited use of commonsense knowledge in high-level applications. We then present a set of vision and reasoning-based methods to solve several applications and show that these approaches benefit in terms of accuracy and interpretability from the explicit use of knowledge and reasoning. We propose novel knowledge representations of image, knowledge acquisition methods, and a new implementation of an efficient probabilistic logical reasoning engine that can utilize publicly available commonsense knowledge to solve applications such as visual question answering, image puzzles. Additionally, we identify the need for new datasets that explicitly require external commonsense knowledge to solve. We propose the new task of Image Riddles, which requires a combination of vision, and reasoning based on ontological knowledge; and we collect a sufficiently large dataset to serve as an ideal testbed for vision and reasoning research. Lastly, we propose end-to-end deep architectures that can combine vision, knowledge and reasoning modules together and achieve large performance boosts over state-of-the-art methods.
Knowledge and reasoning have seen scarce use in image understanding applications. In this thesis, we demonstrate the utilities of incorporating background knowledge and using explicit reasoning in image understanding applications. We first present a comprehensive survey of the previous work that utilized background knowledge and reasoning in understanding images. This survey outlines the limited use of commonsense knowledge in high-level applications. We then present a set of vision and reasoning-based methods to solve several applications and show that these approaches benefit in terms of accuracy and interpretability from the explicit use of knowledge and reasoning. We propose novel knowledge representations of image, knowledge acquisition methods, and a new implementation of an efficient probabilistic logical reasoning engine that can utilize publicly available commonsense knowledge to solve applications such as visual question answering, image puzzles. Additionally, we identify the need for new datasets that explicitly require external commonsense knowledge to solve. We propose the new task of Image Riddles, which requires a combination of vision, and reasoning based on ontological knowledge; and we collect a sufficiently large dataset to serve as an ideal testbed for vision and reasoning research. Lastly, we propose end-to-end deep architectures that can combine vision, knowledge and reasoning modules together and achieve large performance boosts over state-of-the-art methods.
ContributorsAditya, Somak (Author) / Baral, Chitta (Thesis advisor) / Yang, Yezhou (Thesis advisor) / Aloimonos, Yiannis (Committee member) / Lee, Joohyung (Committee member) / Li, Baoxin (Committee member) / Arizona State University (Publisher)
Created2018
Description
Rapid growth of internet and connected devices ranging from cloud systems to internet of things have raised critical concerns for securing these systems. In the recent past, security attacks on different kinds of devices have evolved in terms of complexity and diversity. One of the challenges is establishing secure communication in the network among various devices and systems. Despite being protected with authentication and encryption, the network still needs to be protected against cyber-attacks. For this, the network traffic has to be closely monitored and should detect anomalies and intrusions. Intrusion detection can be categorized as a network traffic classification problem in machine learning. Existing network traffic classification methods require a lot of training and data preprocessing, and this problem is more serious if the dataset size is huge. In addition, the machine learning and deep learning methods that have been used so far were trained on datasets that contain obsolete attacks. In this thesis, these problems are addressed by using ensemble methods applied on an up to date network attacks dataset. Ensemble methods use multiple learning algorithms to get better classification accuracy that could be obtained when the corresponding learning algorithm is applied alone. This dataset for network traffic classification has recent attack scenarios and contains over fifteen attacks. This approach shows that ensemble methods can be used to classify network traffic and detect intrusions with less training times of the model, and lesser pre-processing without feature selection. In addition, this thesis also shows that only with less than ten percent of the total features of input dataset will lead to similar accuracy that is achieved on whole dataset. This can heavily reduce the training times and classification duration in real-time scenarios.
ContributorsPonneganti, Ramu (Author) / Yau, Stephen (Thesis advisor) / Richa, Andrea (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2019
Description
Network mining has been attracting a lot of research attention because of the prevalence of networks. As the world is becoming increasingly connected and correlated, networks arising from inter-dependent application domains are often collected from different sources, forming the so-called multi-sourced networks. Examples of such multi-sourced networks include critical infrastructure networks, multi-platform social networks, cross-domain collaboration networks, and many more. Compared with single-sourced network, multi-sourced networks bear more complex structures and therefore could potentially contain more valuable information.
This thesis proposes a multi-layered HITS (Hyperlink-Induced Topic Search) algorithm to perform the ranking task on multi-sourced networks. Specifically, each node in the network receives an authority score and a hub score for evaluating the value of the node itself and the value of its outgoing links respectively. Based on a recent multi-layered network model, which allows more flexible dependency structure across different sources (i.e., layers), the proposed algorithm leverages both within-layer smoothness and cross-layer consistency. This essentially allows nodes from different layers to be ranked accordingly. The multi-layered HITS is formulated as a regularized optimization problem with non-negative constraint and solved by an iterative update process. Extensive experimental evaluations demonstrate the effectiveness and explainability of the proposed algorithm.
This thesis proposes a multi-layered HITS (Hyperlink-Induced Topic Search) algorithm to perform the ranking task on multi-sourced networks. Specifically, each node in the network receives an authority score and a hub score for evaluating the value of the node itself and the value of its outgoing links respectively. Based on a recent multi-layered network model, which allows more flexible dependency structure across different sources (i.e., layers), the proposed algorithm leverages both within-layer smoothness and cross-layer consistency. This essentially allows nodes from different layers to be ranked accordingly. The multi-layered HITS is formulated as a regularized optimization problem with non-negative constraint and solved by an iterative update process. Extensive experimental evaluations demonstrate the effectiveness and explainability of the proposed algorithm.
ContributorsYu, Haichao (Author) / Tong, Hanghang (Thesis advisor) / He, Jingrui (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2018
Description
This research paper examines Guillaume Lekeu's Sonata for Piano and Violin (1892) from the perspective of a collaborative pianist, providing historical background, an analysis of the work's musical structure, and performance practice insights. Each chapter offers the performer a deeper understanding of various aspects concerning the work, including an in-depth analysis of cyclical features used by Lekeu.
Lekeu was strongly influenced by his teacher, César Franck, and in particular by Franck's use of cyclic techniques, which profoundly impacted Lekeu's Sonata for Piano and Violin. The cyclic treatment, which includes cyclic themes, cyclic motives, and non-cyclic themes is discussed, enabling performers to achieve a relevant structural approach to this work. A performance guide includes practical advice for the interpretation and performance of the work, along with piano pedaling suggestions. The integration of these aspects enables a pianist to gain a better understanding and appreciation of Lekeu's Sonata for Piano and Violin.
Lekeu was strongly influenced by his teacher, César Franck, and in particular by Franck's use of cyclic techniques, which profoundly impacted Lekeu's Sonata for Piano and Violin. The cyclic treatment, which includes cyclic themes, cyclic motives, and non-cyclic themes is discussed, enabling performers to achieve a relevant structural approach to this work. A performance guide includes practical advice for the interpretation and performance of the work, along with piano pedaling suggestions. The integration of these aspects enables a pianist to gain a better understanding and appreciation of Lekeu's Sonata for Piano and Violin.
ContributorsZhang, Dongfang (Author) / Ryan, Russell (Thesis advisor) / Campbell, Andrew (Committee member) / Rogers, Rodney (Committee member) / Arizona State University (Publisher)
Created2018
Description
Louis Hector Berlioz (1803-1869) was a pioneer of 19th century Romanticism in France. In the mid-19th century, he broke the traditional mold by connecting poetry and music through French song. This development transformed French song from the simple and structured Romance of the 18th century into the structural freedom of what he established as the a Mélodie. His song cycle Les nuits d’été, op 7 was composed first for voice and piano in 1841 and later arranged for voice and orchestra in 1856. After the 1856 orchestral version was completed, Les nuits d’été received greater recognition than it had from its original scoring for voice and piano.
This paper examines three major aspects to Les nuits d’été. First, it will discuss the reasons why Berlioz re-scored the work for orchestra and transposed the vocal part for various voice types in this later orchestral version. Second, it examines the difference between musical interactions in these two versions by comparing the existing scores of each version with its respective accompaniment based on Berlioz's use of word painting. Finally, this paper provides the author's original transcription of Les nuits d’été in a version for voice and piano that incorporates the later orchestral versions which were not included in the original version for voice and piano.
This paper examines three major aspects to Les nuits d’été. First, it will discuss the reasons why Berlioz re-scored the work for orchestra and transposed the vocal part for various voice types in this later orchestral version. Second, it examines the difference between musical interactions in these two versions by comparing the existing scores of each version with its respective accompaniment based on Berlioz's use of word painting. Finally, this paper provides the author's original transcription of Les nuits d’été in a version for voice and piano that incorporates the later orchestral versions which were not included in the original version for voice and piano.
ContributorsSeol, Yeojin (Author) / Campbell, Andrew (Thesis advisor) / Britton, David (Committee member) / Ryan, Russell (Committee member) / Rogers, Rodney (Committee member) / Arizona State University (Publisher)
Created2017