A
Andreas Spanias
Researcher at Arizona State University
Publications - 512
Citations - 8918
Andreas Spanias is an academic researcher from Arizona State University. The author has contributed to research in topics: Speech coding & Speech processing. The author has an hindex of 36, co-authored 490 publications receiving 7895 citations. Previous affiliations of Andreas Spanias include Arizona's Public Universities & Intel.
Papers
More filters
Proceedings ArticleDOI
Development of mobile sensing apps for DSP applications
TL;DR: The enhanced sensing capabilities of modern mobile devices have been explored by developing DSP tools that allow interfaces to on-board and external sensors.
Proceedings ArticleDOI
Removing data with noisy responses in regression analysis
TL;DR: An empirically estimable bound on the regression error based on a Euclidean minimum spanning tree generated from the data is derived and an iterative approach to remove data with noisy responses from the training set is proposed.
Re-sonification of geographic sound activity using acoustic, semantic, and social information
Alex Fink,Brandon Mechtley,Gordon Wichern,Jinru Liu,Harvey Thornburg,Andreas Spanias,Grisha Coleman +6 more
TL;DR: This work has developed an ontological framework to determine the importance of sound and concepts to one another using acoustic, semantic, and social information and uses this framework in the automated design of a generative soundscape model purposed to identify and re-sonify sounds that impart relevant information about a geographic region.
Proceedings ArticleDOI
Web-based experiments for introducing speech recognition basics in a DSP course
TL;DR: In this paper, the authors describe Web-based educational software tools tailored to expose students in an undergraduate DSP course to the basics of hidden Markov model (HMM)-based speech recognition.
Posted Content
Attention Models with Random Features for Multi-layered Graph Embeddings.
TL;DR: This paper proposes to use attention models for effective feature learning, and develops two novel architectures that exploit the inter-layer dependencies for building multi-layered graph embeddings, and shows that using simple random features is an effective choice, even in cases where explicit node attributes are not available.