Select a calendar:
Filter December Events by Event Type:
SUNMONTUEWEDTHUFRISAT
University Calendar
Events for December
-
PhD Defense - Ross Mead
Thu, Dec 03, 2015 @ 02:00 PM - 04:00 PM
Thomas Lord Department of Computer Science
University Calendar
PhD Defense - Ross Mead
Title: Situated Proxemics and Multimodal Communication: Space, Speech, and Gesture in Human-Robot Interaction
Committee: Maja Mataric, Gaurav Sukhatme, Gisele Ragusa (external member)
Abstract:
To facilitate face-to-face human-robot interaction (HRI), a sociable robot must employ multimodal communication mechanisms similar to those used by humans: speech production (via speakers), speech recognition (via microphones), gesture production (via physical embodiment), and gesture recognition (via cameras or motion trackers). Like any other signals, these social signals are affected by distance and interference present in the medium through which they travel. People often compensate for this attenuation by adjusting the production of their social signals to compensate for these effects-”for example, by speaking louder, using more broad gestures, or moving closer. How can a sociable robot do the same?
This dissertation investigates how social (speech and gesture) and environmental (loud noises and reduced visibility) factors influence positioning and communication between humans and sociable robots. Specifically, this research answers the following questions: 1) How should a robot dynamically adjust its position (proxemics) to maximize its automated recognition of human social signals? 2) How should a robot adjust its own communication behaviors to maximize human perceptions of its social signals? 3) How can a robot quickly adapt its models of proxemic and communication behavior to differences in human social signal perception?
This research formalizes an extensible unifying framework for situated proxemics and multimodal communication in HRI. The framework considers how both humans and robots experience social signals in face-to-face interactions. Data collections were conducted to inform probabilistic graphical models based on the framework that predict how speech and gesture are produced (transmitted) and perceived (received) by both humans and robots at different distances and under environmental interference.
This work integrates the resulting data-driven models into autonomous proxemic behavior and multimodal communication control system for sociable robots. The robot control system selects positioning parameters to maximize its ability to automatically recognize natural human speech and gestures. Furthermore, the robot control system can dynamically adjust its own speech and gestures to maximize human perceptions of its social signals. Experiments were conducted that successfully evaluated user acceptance of the autonomous robot proxemic control system, demonstrating that human users are willing to adapt their behavior preferences in exchange for improved robot performance in social contexts.
This research establishes a foundational component of HRI, enabling the development of robust controllers for socially intelligent robots in complex environments.
Furthermore, this work has implications for technology personalization in socially assistive contexts with people with special needs, such as older adults, children with autism spectrum disorders, and people with hearing or visual impairments or sensitivities.
Location: Ronald Tutor Hall of Engineering (RTH) - 406
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
Code Dojo: K-12 Schools & Families Connect with USC for
Tue, Dec 08, 2015 @ 09:00 AM - 04:00 PM
USC Viterbi School of Engineering, Viterbi School of Engineering K-12 STEM Center
University Calendar
In conjunction with CS Education Week & the global Hour of Code movement, Code Dojo supports K-12 teachers and families learning to code. Meet on Google Hangouts (ViterbiOutreach) or schedule a session in the USC Community Computing Center. http://www.viterbi.usc.edu/k-12/coding/code-dojo.htm
VAST: Viterbi Adopt-a-Student, Adopt-a-Teacher.More Information: Code Dojo Poster.pdf
Location: USC Community Computing Center
Audiences: Students, Schools, and Families
Contact: Katie Mills
-
PhD Defense - Sangwon Lee
Fri, Dec 11, 2015 @ 03:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title: A Framework for Runtime Energy Efficient Mobile Execution
Ph.D. Candidate: Sangwon Lee
Friday, December 11, 2015
3:00PM
EEB 349
Abstract:
As mobile applications become more complex, computation offloading on mobile devices emerges for one possible energy saving approach as well as improving performance. Existing approaches to offload requires significant amount of method call state to be transferred to the remote server. Furthermore, existing approaches are unable to handle the increasingly popular native method calls that are embedded in most mobile applications.
In this thesis we present a framework named FREEME for efficient offloading of computations to a remote server. At the first, we present a comprehensive quantification of a mobile phone's energy consumption using an in-filed deployed Wireless Body Are Networks called KNOWME. We quantify the energy impact of different programming paradigms, sensing modalities, data storage, and conflicting computation and communication demands. Based on the knowledge gained from the measurement studies, we propose an Active Energy Profiling strategy that uses short profiling periods to automatically determine the most energy efficient choices for running a WBAN. After that, we propose propose a novel static analyzer to identify offloadable methods from a legacy Android application.
The proposed analyzer provides a comprehensive analysis that can analyze both java method and native method.
At the last, we present the FREEME in detail; (1) it automatically analyzes the Java class methods and user-defined native methods in Android applications to identify target methods for remote execution. (2) FREEME's static analysis identifies minimum set of data elements that are necessary for remote execution thereby shrinking the size of data transferred to the server. The server also optimizes the amount of data it sends back to the mobile phone by eliminating data transfers of unmodified data. FREEME implements these approaches within the Android framework by developing novel static analysis and object serialization approaches. We evaluated FREEME on Android phones and show that significant energy and latency reductions can be achieved with FREEME.
Biography:
Sangwon Lee is a Ph.D. candidate in the department of computer science at the University of Southern California. In 2008, he was working in LG Electronics as a Senior Research Engineer. He received the M.S. degree in computer science from from the University of Southern California, and the B.A. degree in computer science from the Seoul National University of Technology, Seoul, South Korea. Before his studies at USC, he worked as a system architecture and a DBA for 6 years. He established his own company, Interrush Korea Inc., in 2002. His general interest is in mobile applications and wireless sensor networks.
Committee:
Prof. Murali Annavaram (chair)
Prof. Bhaskar Krishnamachari
Prof. Aiichiro Nakano
Location: 349
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
PhD Defense - Elnaz Nouri
Wed, Dec 23, 2015 @ 10:30 AM - 12:30 PM
Thomas Lord Department of Computer Science
University Calendar
Title: Cultural Computational Agents
PhD Candidate: Elnaz Nouri
Time: Wednesday, December 23, 2015 10:30 AM PDT
Place: SAL 213
People's cultural background influences how they make decisions. Faced with the same set of choices in the same tasks, people from different cultures tend to make different decisions. Cultural variations in behavior play an influential role in how interpersonal interactions unfold. Therefore, computational agents designed for interacting with people in social interactions (such as virtual humans that are used for training and negotiation) need to be sensitive to culture in order to be effective. The majority of the existing computational decision making models do not account for the observed cultural variations in behavior. This thesis proposes how models of culture-specific decision making can be created using information about the values of people from a culture, and that thses models can be used to support the decision making of computational agents in simulating behavior of people from a specific culture. The Multi Attribute Relational Values model (called MARV) assumes that decisions are evaluated based on a set of social and relational considerations. Different approaches for developing the decision making models within the MARV framework based on the availability of different sources of information on the culture are explored and compared with one another. An online interactive web framework is implemented that allows inter-cultural and intra-cultural behavioral data among users and computational agents collection through crowdsourcing platforms. It is shown that the culture-sensitive decision making models enable the virtual agents to simulate, predict and appropriately respond to the behavior of people from different cultures.
Committee:
Prof. David Traum (chair)
Prof. Paul Rosenbloom
Prof. Morteza Dehghani
Prof. Jerry Hobbs
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
