Matching Items (91)
Filtering by
- Genre: Masters Thesis
- Creators: Yang, Yezhou
- Member of: ASU Electronic Theses and Dissertations
Description
This composition was commissioned by the Orgelpark to be performed in Amsterdam in September 2011 during Gaudeamus Muziekweek. It will be performed by the vocal group VocaalLab Nederland. It is scored for four vocalists, organ, tanpura, and electronic sound. The work is a culmination of my studies in South Indian Carnatic rhythm, North Indian classical singing, and American minimalism. It is a meditation on the idea that the drone and pulse are micro/macro aspects of the same phenomenon of vibration. Cycles are created on the macroscale through a mathematically defined scale of harmonic/pitch relationships. Cycles are created on the microscale through the subdivision and addition of rhythmic pulses.
ContributorsAdler, Jacob (Composer) / Rockmaker, Jody (Thesis advisor) / Feisst, Sabine (Committee member) / Etezady, Roshanne, 1973- (Committee member) / Arizona State University (Publisher)
Created2011
Description
Delirium is a piece for large wind ensemble that synthesizes compositional techniques to generate unique juxtapositions of contrasting musical elements. The piece is about 8:30 long and uses the full complement of winds, brass, and percussion. Although the composition begins tonally, chromatic alterations gradually shift the melodic content outside of the tonal center. In addition to changes in the melody, octatonic, chromatic, and synthetic scales and quartal and quintal harmonies are progressively introduced throughout the piece to add color and create dissonance. Delirium contains four primary sections that are all related by chromatic mediant. The subdivisions of the first part create abrupt transitions between contrasting material, evocative of the symptoms of delirium. As each sub-section progresses, the A minor tonality of the opening gradually gives way to increased chromaticism and dissonance. The next area transitions to C minor and begins to feature octatonic scales, secundal harmonies, and chromatic flourishes more prominently. The full sound of the ensemble then drops to solo instruments in the third section, now in G# minor, where the elements of the previous section are built upon with the addition of synthetic scales and quartal harmonies. The last division, before the recapitulation of the opening material, provides a drastic change in atmosphere as the chromatic elements from before are removed and the tense sound of the quartal harmonies are replaced with quintal sonorities and a more tonal melody. The tonality of this final section is used to return to the opening material. After an incomplete recapitulation, the descending motive that is used throughout the piece, which can be found in measure 61 in the flutes, is inverted and layered by minor 3rds. This inverted figure builds to the same sonority found in measure138, before ending on an F# chord, a minor third away from the A minor tonal center of the opening and where the piece seems like it should end.
ContributorsBell, Jeremy, 1986- (Composer) / Rogers, Rodney (Thesis advisor) / Oldani, Robert (Committee member) / Levy, Benjamin (Committee member) / Arizona State University (Publisher)
Created2011
Description
Throughout history composers and artists have been inspired by the natural world. Nature's influence on music is extraordinary, though water in particular, has had a unique magnetic pull. The large number of compositions dealing with water, from Handel's Water Music (1717) to Ros Bandt's and Leah Barclay's Rivers Talk (2012), reflects this continuous fascination. Since the late 1940s, composers have ventured further and brought actual sounds from the environment, including water recorded on tape, into the musical arena. Moreover, since the 1960s, some composers have nudged their listeners to become more ecologically aware. Much skepticism exists, as with any unconventional idea in history, and as a result compositions belonging to this realm of musique concrète are not as widely recognized and examined as they should be. In this thesis, I consider works of three composers: Annea Lockwood, Eve Beglarian, and Leah Barclay, who not only draw inspiration from nature, but also use their creativity to call attention to pristine environments. All three composers embrace the idea that music can be broadly defined and use technology as a tool to communicate their artistic visions. These artists are from three different countries and represent three generations of composers who set precedents for a new way of composing, listening to, performing, and thinking about music and the environment. This thesis presents case studies of Lockwood's A Sound Map of the Danube River, Beglarian's Mississippi River Project, and Barclay's Sound Mirrors. This thesis draws on unpublished correspondence with the composers, analytical theories of R. Murray Schafer, Barry Truax, and Martijn Voorvelt, among others, musicological publications, eco-critical and environmental studies by Al Gore, Bill McKibben, and Vandana Shiva, as well as research by feminist scholars. As there is little written on music and nature from an eco-critical and eco-feminist standpoint, this thesis will contribute to the recognition of significant figures in contemporary music that might otherwise be overlooked. In this study I maintain that composers and sound artists engage with sounds in ways that reveal aspects of particular places, and their attitudes toward these places to lead listeners toward a greater ecological awareness.
ContributorsRichardson, Jamilyn (Author) / Feisst, Sabine (Thesis advisor) / Solís, Ted (Committee member) / Norton, Kay (Committee member) / Arizona State University (Publisher)
Created2012
Description
Three Meditations on the Philosophy of Boethius is a musical piece for guitar, piano interior, and computer. Each of the three movements, or meditations, reflects one level of music according to the medieval philosopher Boethius: Musica Mundana, Musica Humana, and Musica Instrumentalis. From spatial aspects, through the human element, to letting sound evolve freely, different movements revolve around different sounds and sound producing techniques.
ContributorsDori, Gil (Contributor) / Hackbarth, Glenn (Thesis advisor) / DeMars, James (Committee member) / Feisst, Sabine (Committee member) / Arizona State University (Publisher)
Created2013
Description
In the 1930s, with the rise of Nazism, many artists in Europe had to flee their homelands and sought refuge in the United States. Austrian composer Hanns Eisler who had risen to prominence as a significant composer during the Weimar era was among them. A Jew, an ardent Marxist and composer devoted to musical modernism, he had established himself as a writer of film music and Kampflieder, fighting songs, for the European workers' movement. After two visits of the United States in the mid-1930s, Eisler settled in America where he spent a decade (1938-1948), composed a considerable number of musical works, including important film scores, instrumental music and songs, and, in collaboration with Theodor W. Adorno, penned the influential treatise Composing for the Films. Yet despite his substantial contributions to American culture American scholarship on Eisler has remained sparse, perhaps due to his reputation as the "Karl Marx in Music." In this study I examine Eisler's American exile and argue that Eisler, through his roles as a musician and a teacher, actively sought to enrich American culture. I will present background for his exile years, a detailed overview of his American career as well as analyses and close readings of several of his American works, including three of his American film scores, Pete Roleum and His Cousins (1939), Hangmen Also Die (1943), and None But the Lonely Heart (1944), and the String Quartet (1940), Third Piano Sonata (1943), Woodbury Liederbüchlein (1941), and Hollywood Songbook (1942-7). This thesis builds upon unpublished correspondence and documents available only in special collections at the University of Southern California (USC), as well as film scores in archives at USC and the University of California, Los Angeles. It also draws on Eisler studies by such European scholars as Albrecht Betz, Jürgen Schebera, and Horst Weber, as well as on research of film music scholars Sally Bick and Claudia Gorbman. As there is little written on the particulars of Eisler's American years, this thesis presents new facts and new perspectives and aims at a better understanding of the artistic achievements of this composer.
ContributorsBoyd, Caleb (Author) / Feisst, Sabine (Thesis advisor) / Levy, Benjamin (Committee member) / Oldani, Robert (Committee member) / Arizona State University (Publisher)
Created2013
Description
Gloria is a work written for SATB choir and brass quintet that uses the traditional Latin text of the Gloria found in the ordinary of the Mass. The piece is approximately fourteen minutes and explores a variety of textures, colors, and timbres of the brass quintet and choir. The composition uses quartal sonorities mixed with upper tertian structures while avoiding simple triads and stable root position voicings until the most important climactic moments. The Gloria opens with a fanfare presenting the initial rhythmic motive in a call and response between the brass and choir before the irregular meters of the A section enter. The piece develops a variety of sonorities, pitch collections, and timbres before arriving at the first climactic moment on the text "Rex" (King). The music slowly comes to a point of repose with a brass interlude revealing the motives used in the B section. The choir begins the B section a cappella on the text "Dómine Fili unigénite, Jésu Chríste" (Lord Jesus Christ, the only begotten Son). The section features a dialogue between the brass and choir, though the two groups never sound together. The section includes a lyrical soprano duet incorporating dissonant intervals preceding the choir's response on the text requesting the mercy of the Lord. The section comes to a somber, penitential rest ending with the brass quintet response. The piece gradually builds and accelerates to the second climactic moment on the word "Jésu." From there it once again gains momentum toward the return of the A section on the text "Cum Sáncto Spíritu" (With the Holy Spirit). After a climactic "Amen" section, the composition concludes with a return to the material found in the introduction followed by an affirming brass postlude.
ContributorsRichard, Nathan Daniel (Author) / Rogers, Rodney (Thesis advisor) / DeMars, James (Committee member) / Gentry, Gregory (Committee member) / Arizona State University (Publisher)
Created2012
Description
Feature embeddings differ from raw features in the sense that the former obey certain properties like notion of similarity/dissimilarity in it's embedding space. word2vec is a preeminent example in this direction, where the similarity in the embedding space is measured in terms of the cosine similarity. Such language embedding models have seen numerous applications in both language and vision community as they capture the information in the modality (English language) efficiently. Inspired by these language models, this work focuses on learning embedding spaces for two visual computing tasks, 1. Image Hashing 2. Zero Shot Learning. The training set was used to learn embedding spaces over which similarity/dissimilarity is measured using several distance metrics like hamming / euclidean / cosine distances. While the above-mentioned language models learn generic word embeddings, in this work task specific embeddings were learnt which can be used for Image Retrieval and Classification separately.
Image Hashing is the task of mapping images to binary codes such that some notion of user-defined similarity is preserved. The first part of this work focuses on designing a new framework that uses the hash-tags associated with web images to learn the binary codes. Such codes can be used in several applications like Image Retrieval and Image Classification. Further, this framework requires no labelled data, leaving it very inexpensive. Results show that the proposed approach surpasses the state-of-art approaches by a significant margin.
Zero-shot classification is the task of classifying the test sample into a new class which was not seen during training. This is possible by establishing a relationship between the training and the testing classes using auxiliary information. In the second part of this thesis, a framework is designed that trains using the handcrafted attribute vectors and word vectors but doesn’t require the expensive attribute vectors during test time. More specifically, an intermediate space is learnt between the word vector space and the image feature space using the hand-crafted attribute vectors. Preliminary results on two zero-shot classification datasets show that this is a promising direction to explore.
Image Hashing is the task of mapping images to binary codes such that some notion of user-defined similarity is preserved. The first part of this work focuses on designing a new framework that uses the hash-tags associated with web images to learn the binary codes. Such codes can be used in several applications like Image Retrieval and Image Classification. Further, this framework requires no labelled data, leaving it very inexpensive. Results show that the proposed approach surpasses the state-of-art approaches by a significant margin.
Zero-shot classification is the task of classifying the test sample into a new class which was not seen during training. This is possible by establishing a relationship between the training and the testing classes using auxiliary information. In the second part of this thesis, a framework is designed that trains using the handcrafted attribute vectors and word vectors but doesn’t require the expensive attribute vectors during test time. More specifically, an intermediate space is learnt between the word vector space and the image feature space using the hand-crafted attribute vectors. Preliminary results on two zero-shot classification datasets show that this is a promising direction to explore.
ContributorsGattupalli, Jaya Vijetha (Author) / Li, Baoxin (Thesis advisor) / Yang, Yezhou (Committee member) / Venkateswara, Hemanth (Committee member) / Arizona State University (Publisher)
Created2019
Description
Traditional Reinforcement Learning (RL) assumes to learn policies with respect to reward available from the environment but sometimes learning in a complex domain requires wisdom which comes from a wide range of experience. In behavior based robotics, it is observed that a complex behavior can be described by a combination of simpler behaviors. It is tempting to apply similar idea such that simpler behaviors can be combined in a meaningful way to tailor the complex combination. Such an approach would enable faster learning and modular design of behaviors. Complex behaviors can be combined with other behaviors to create even more advanced behaviors resulting in a rich set of possibilities. Similar to RL, combined behavior can keep evolving by interacting with the environment. The requirement of this method is to specify a reasonable set of simple behaviors. In this research, I present an algorithm that aims at combining behavior such that the resulting behavior has characteristics of each individual behavior. This approach has been inspired by behavior based robotics, such as the subsumption architecture and motor schema-based design. The combination algorithm outputs n weights to combine behaviors linearly. The weights are state dependent and change dynamically at every step in an episode. This idea is tested on discrete and continuous environments like OpenAI’s “Lunar Lander” and “Biped Walker”. Results are compared with related domains like Multi-objective RL, Hierarchical RL, Transfer learning, and basic RL. It is observed that the combination of behaviors is a novel way of learning which helps the agent achieve required characteristics. A combination is learned for a given state and so the agent is able to learn faster in an efficient manner compared to other similar approaches. Agent beautifully demonstrates characteristics of multiple behaviors which helps the agent to learn and adapt to the environment. Future directions are also suggested as possible extensions to this research.
ContributorsVora, Kevin Jatin (Author) / Zhang, Yu (Thesis advisor) / Yang, Yezhou (Committee member) / Praharaj, Sarbeswar (Committee member) / Arizona State University (Publisher)
Created2021
Description
Machine learning models can pick up biases and spurious correlations from training data and projects and amplify these biases during inference, thus posing significant challenges in real-world settings. One approach to mitigating this is a class of methods that can identify filter out bias-inducing samples from the training datasets to force models to avoid being exposed to biases. However, the filtering leads to a considerable wastage of resources as most of the dataset created is discarded as biased. This work deals with avoiding the wastage of resources by identifying and quantifying the biases. I further elaborate on the implications of dataset filtering on robustness (to adversarial attacks) and generalization (to out-of-distribution samples). The findings suggest that while dataset filtering does help to improve OOD(Out-Of-Distribution) generalization, it has a significant negative impact on robustness to adversarial attacks. It also shows that transforming bias-inducing samples into adversarial samples (instead of eliminating them from the dataset) can significantly boost robustness without sacrificing generalization.
ContributorsSachdeva, Bhavdeep Singh (Author) / Baral, Chitta (Thesis advisor) / Liu, Huan (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2021
Description
There has been an explosion in the amount of data on the internet because of modern technology – especially image data – as a consequence of an exponential growth in the number of cameras existing in the world right now; from more extensive surveillance camera systems to billions of people walking around with smartphones in their pockets that come with built-in cameras. With this sudden increase in the accessibility of cameras, most of the data that is getting captured through these devices is ending up on the internet. Researchers soon took leverage of this data by creating large-scale datasets. However, generating a dataset – let alone a large-scale one – requires a lot of man-hours. This work presents an algorithm that makes use of optical flow and feature matching, along with utilizing localization outputs from a Mask R-CNN, to generate large-scale vehicle datasets without much human supervision. Additionally, this work proposes a novel multi-view vehicle dataset (MVVdb) of 500 vehicles which is also generated using the aforementioned algorithm.There are various research problems in computer vision that can leverage a multi-view dataset, e.g., 3D pose estimation, and 3D object detection. On the other hand, a multi-view vehicle dataset can be used for a 2D image to 3D shape prediction, generation of 3D vehicle models, and even a more robust vehicle make and model recognition. In this work, a ResNet is trained on the multi-view vehicle dataset to perform vehicle re-identification, which is fundamentally similar to a vehicle make and recognition problem – also showcasing the usability of the MVVdb dataset.
ContributorsGuha, Anubhab (Author) / Yang, Yezhou (Thesis advisor) / Lu, Duo (Committee member) / Banerjee, Ayan (Committee member) / Arizona State University (Publisher)
Created2022