Project Name
Virtual Sprouts

Project Description
The Virtual Sprouts project is an interdisciplinary effort to build a highly engaging, mobile game (in Unity3D on Android or iPhone) to improve middle school and the community’s understanding of health science advances in obesity and nutrition. ICT partners include the Keck School of Medicine, the School of Cinema and Television, and the Rossier School of Education. The system uses modern gaming technology and intelligent tutoring techniques to help entertain and educate children about healthy eating, cooking, and choices.

Job Description
The Virtual Sprouts intern will work closely with an ICT programmer, the project leader (Lane), and other partners on the team (many of whom are students in the departments listed above) to support development of the game as well as a pedagogical agent that will inhabit the game. This will include design, implementation, and testing in Java and Unity3D.

Preferred Skills

• Basic (or better) knowledge of Unity 3D
• Introductory (or better) knowledge of artificial intelligence
• Some knowledge of Eclipse or similar IDE

Project Name
COSMOS Social Simulation Lab

Project Description
The COSMOS Social Simulation Lab works on the modeling and simulation of social systems from small group to societal level interactions and approaches to validating these models. Our approach to simulation relies on multi-agent techniques where autonomous, goal driven agents are used to model the entities in the simulation, including individuals, groups, organizations and structures. A key application area for these models is their use in serious games where players interact with the entities in the simulation.

Job Description
The research assistant will investigate approaches to creating computational models of players. These models can be used to replace players in a simulation as well as to automate the play testing of games. The key part of the task will involve using machine learning techniques to derive models of players from logs of their actions in a simulation. This may involve analyzing our existing data sets of player interactions and/or also creating a testbed for collecting new player data from existing simulation-based games.

Preferred Skills

• Knowledge of multi-agent systems and social simulation
• Knowledge of machine learning techniques
• Experience with python programming

Project Name
MxR Lab: User Interface

Project Description
This project involves designing and developing novel user interfaces that incorporate body motion and gesture. The goal is to provide intuitive authoring tools that can be used by non-programmers to create real-time interactions for body-based user interfaces such as the Microsoft Kinect.

Job Description
Duties will involve designing and developing graphical user interfaces that will enable users to create gestural interactions using ICT’s Flexible Action and Articulated Skeleton Toolkit (FAAST).

Preferred Skills

• Familiarity with HCI and principles of user interface design
• Programming in C++ or similar languages
• Experience with motion sensing technology such as the Microsoft Kinect

Project Name
Mixed Reality Research and Development Stretching Space

Project Description
This project involves developing and testing novel perceptual illusions that enable users to walk through expansive virtual reality environments in small physical spaces.

Job Description
Duties will involve brainstorming and rapid prototyping of novel VR interaction/locomotion techniques, developing virtual environments using the Unity3D game engine, running user studies, and analyzing experiment data.

Preferred Skills

• Development experience using game engines such as Unity3D
• Prior experience with virtual reality technology (e.g. head-mounted displays, motion tracking)
• Programming in C++, C#, or similar languages
• Familiar with experimental design and user study procedures

Project Name
MxR Lab: XData Visualization

Project Description
This project involves developing and evaluating interactive visualization techniques for large data sets, particularly focused on tablet and multi-touch displays.

Job Description
Duties will involve development of new touch and body-based interaction techniques, programming with Java and the Unity3D game engine, designing and running user studies, and analyzing experiment data.

 Preferred Skills: 
- Development experience using game engines such as Unity3D
- Prior experience with visualization
- Prior experience with virtual reality technologies (e.g. head-mounted displays, motion tracking)
- Programming in C#, Java, Objective C or similar languages
- Familiar with experimental design and user study procedures is a plus
- Prior experience with mobile development is a plus


Preferred Skills

• Development experience using game engines such as Unity3D
• Prior experience with visualization
• Prior experience with virtual reality technologies (e.g. head-mounted displays, motion tracking)

• Programming in C#, Java, Objective C or similar languages

• Familiar with experimental design and user study procedures is a plus
• Prior experience with mobile development is a plus

Project Name
Mixed Reality Research and Development Sharing Space

Project Description
This project involves developing and evaluating display techniques for presenting virtual humans in mixed reality and virtual reality.

Job Description
Duties will involve programming virtual characters using the Unity3D game engine and Smartbody, designing and running user studies, and analyzing experiment data.

Preferred Skills: 
- Development experience using game engines such as Unity3D
- Prior experience with virtual reality technology (e.g. head-mounted displays, motion tracking)
.

Preferred Skills

• Programming C++, C#, or similar languages
• Familiar with experimental design and user study procedures

• Experience with audio programming and design is a plus

Project Name
SmartBody: A real-time character animation system

Project Description
Develop and improve the capabilities of a real-time 3D animation system. The animation system, SmartBody, supports a wide range of behavior for animation characters, including: locomotion, steering, speech, object interaction, nonverbal behavior such as gesturing and head nodding, emotional expression, and gazing, among others.

We seek to improve upon these capabilities by adding new character behaviors, bolstering the robustness of our present behaviors, and improving the performance and uses of each behavior. Improvements will involve development of new algorithms as well as implementation of existing animation algorithms from recent research. Areas of development include: locomotion, motion planning, facial animation, physics, automated hand postures, nonverbal behavior, crowd behavior, planning or perhaps other related to area of interest and expertise of the intern.

Job Description
Develop algorithms to produce new or improve animation behavior. Algorithms will be written in C++. Candidates should be aware of animation research, fluent in C++ programming, understand OpenGL. Familiarity with Python language is helpful. Knowledge of multi-platform development is also helpful (osx, linux, windows, ios, android).

Prefer PhD student in area of computer graphics/animation who is working on one or more areas described above.

Preferred Skills

• C++
• OpenGL
• Knowledge of current animation research

Project Name
Creating the Avatar Investment Metric (AIM)

Project Description
AIM is part of a systematic research effort looking at the psychological, emotional and social effects of personal avatar usage, with the goal of developing a widely distributable Avatar Investment Metric (AIM). We begin with a pilot study looking at factors impacting users’ initial avatar selection to establish the psychological characteristics expressed in selecting an avatar. This information will be used in the design of the initial AIM that we will then deploy into virtual world populations for further refinement. We will then correlate the refined AIM with both physiological measures (in partnership with the National Intrepid Center in Bethesda, MD) and with neurological measures (in collaboration with UCSD Neurosciences expert Dr. Todd Coleman) to leverage the most recent expertise in non-invasive body signal data collection and analysis.

Job Description
Research Assistant to work on studies about avatar usage. The applicant will have experience in supporting studies about usability or the psychology domain.

Preferred Skills

• Experience with Human Subject studies
• Knowledge of statistics to analyze data from studies
• Experience with MMOG games and/or virtual worlds a plus

Project Name
Avatar Investment Metric (AIM)

Project Description
AIM is part of a systematic research effort looking at the psychological, emotional and social effects of personal avatar usage, with the goal of developing a widely distributable Avatar Investment Metric (AIM). We begin with a pilot study looking at factors impacting users’ initial avatar selection to establish the psychological characteristics expressed in selecting an avatar. This information will be used in the design of the initial AIM that we will then deploy into virtual world populations for further refinement. We will then correlate the refined AIM with both physiological measures (in partnership with the National Intrepid Center in Bethesda, MD) and with neurological measures (in collaboration with UCSD Neurosciences expert Dr. Todd Coleman) to leverage the most recent expertise in non-invasive body signal data collection and analysis.

Job Description
This position is to work with researchers to create graphics, models and animations in support of the development of the Avatar Investment Metric, an instrument that will be developed and used toy measure how closely a person identifies with their avatar in a virtual world. The applicant will contribute to building in-world scenarios to support pilto studies of avatars within virtual worlds.

Preferred Skills

• Experience with 2D and 3D asset creation fro games or virtual worlds
• Proficient in art processes, including drawing, story boarding and or texture map creation
• Ability to work well in groups
• Passion for research into social media, MMOPG games and virtual worlds.

Project Name
Computational Modeling of Human Communication

Project Description
Human communication involves continuous generation and interpretation of signals from several channels. Words, prosody, gestures and facial expressions are typically coordinated seamlessly among interaction participants, often resulting in a communicative total that is more than the sum of its parts. This project aims to model this potentially complex interaction of verbal and non-verbal information in human communication, and how it conveys intended messages or elicits emotions, using data-driven computational approaches.

Job Description
The intern will analyze human communication data (e.g. dialogue transcripts, videos) and work closely with faculty on computational modeling of how verbal (words) and non-verbal (gestures, facial expressions) information contribute to communicative intent (meaning, dialogue acts, sentiment, etc). This may involve collecting human judgments using Amazon Mechanical Turk, or building data-driven models from existing annotation or standard datasets, or designing new representations for verbal and/or non-verbal information in computational models.

Preferred Skills

• Familiarity with one or more of: natural language processing, vision, audio/video processing
• Solid programming skills
• Desired: some experience with machine learning or human communication
• Desired: some research experience

Project Name
Telepresence Rehabilitation

Project Description
This project builds upon our previous work with the Microsoft Kinect and our own specifically tailored software for rehabilitation. Video game technologies may provide a means to create rehabilitation and exercise environments that will allow health care professionals to precisely deliver and control complex dynamic, interactive, three dimensional stimulus presentations.

Job Description
We are looking for a motivated research assistant to take part in a series of data collection sessions over the summer period. We will be working with Clinicians and patient populations around the Los Angeles area using our Microsoft Kinect game-based rehabilitation tool. Data collection will require setting up the technology, obtaining consent from participants, observing interactions, interviewing participants and analyzing data.

Preferred Skills

• Experience with data collection and analysis
• Experience with user testing and interviewing techniques
• Clinical and/or rehabilitation experience a plus

Project Name
A High-Fedelity Morphable Model Enabling “Photos to Faces”

Project Description
Extracting complete 3D models of human faces from collections of uncontrolled photographs has been a goal in computer graphics and computer vision for some time. Methods have been developed that employ morphable models or facial feature detection with some degree of success, but the extracted models lack realism especially when viewed under novel illumination or in closer detail than the source photographs, and typically cannot be animated. We plan to extend the concept of morphable face models to include multiple layers of information, including the highest resolution pore detail, in order to obtain more realistic reconstructions of faces from photographs in uncontrolled conditions, and to serve as a high-detail statistical model of the human face in other ICT projects.

Job Description
This internship will focus on employing optical flow and facial feature detection algorithms to detect and localize faces in photographs from multiple sources. As the project progresses, focus will shift towards corresponding facial features between faces and developing facial detail texture synthesis algorithms.

Preferred Skills

• C++, OpenGL, GPU programming
• Experience with computer vision techniques: multi-camera stereo, optical flow, facial feature, detection, bilinear morphable models, texture synthesis, markov random fields
• Operating System: Windows

Project Name
Evaluation of Virtual Character Authoring Tools

Project Description
We are currently developing a new information visualization approach for authoring Virtual Human (VH) knowledge. Our system represents parts of the VH knowledge database as objects in 3D space and arranges the objects according to the relationships that exist in the data. The central idea is that a correct knowledge base will be shown as a well-formed regular spatial structure, while database errors would be depicted as irregularities in the visualization. Our hypothesis is that such a visualization will provide a VH designer with a concise and informative summary of the knowledge database. It will help the designer to manage the database and maintain its integrity. Ultimately, it will allow the designer to build better, more robust VHs.

Job Description
The intern will help the project supervisor to conduct a user study of the new authoring system that uses the information visualization approach. The intern will help with recruiting the subjects, train them to use the system, administer study questionnaires, and supervise the subjects using the system. The intern will also collect and analyze the experimental data.

Preferred Skills

• Good spoken and written English (native or near-native competence preferred) to deliver instructions to the subjects
• Ability to work in a collaborative environment as well as work independently
• Some knowledge of scientific experimental method (e.g. basic statistics)

Project Name
Development of Interactive Spoken Dialogue Games

Project Description
This project will build on existing software infrastructure at ICT to implement a fun, fast-paced interactive spoken dialogue game that can be played with a virtual human.

In this project, the intern will develop software to implement game mechanics for a collaborative game, similar to a TV game show, where the players converse rapidly and score points as they complete a series of challenges displayed on a computer screen.

The game can be played by two human players working together, or by a single human player working with a virtual human partner. In this project, ICT’s existing virtual human architecture and dialogue system pipeline will be leveraged to implement gameplay between human players and virtual humans.

This project involves the design and implementation of game mechanics, configuration of the virtual human player, and implementation of data capture capabilities. The real-time speech of both players will be
captured as the game progresses, creating a corpus of spoken dialogue data for use in dialogue systems research. The intern will assess their design and implementation by collecting a small corpus of spoken dialogue games using the software.

In addition to gaining experience working with virtual humans and ICT’s virtual human architecture, the intern will gain exposure to a number of issues in the collection and analysis of natural language data.

Job Description
The successful intern will develop software to implement game mechanics and data capture for a fast-paced interactive spoken dialogue game, and will collect a data set using this implementation.

Preferred Skills

• Required: Master’s or PhD student in computer science
• Required: Strong programming ability and experience
• Desired: Java programming experience
• Desired: Some familiarity with dialogue systems or natural language dialogue

Project Name
Incremental Speech Processing Techniques for Fast-Paced Dialogue Games

Project Description
This research project will develop new incremental speech processing techniques for use in dialogue systems that play fun, fast-paced interactive dialogue games with human users.

Human-human conversation is highly interactive, on a very rapid time-scale, with participants providing a range of rapid and overlapping responses, as speech is occurring, to facilitate an efficient communication process. In comparison, state of the art spoken dialogue systems are much less interactive: they provide little or no feedback to users as they are speaking, and they often enforce a strict turn-taking format that eliminates overlapping speech altogether.

Building better spoken dialogue systems that are more interactive and more human-like in how they respond to user speech requires new computational techniques that can understand and respond to speech incrementally, word by word, as a user is speaking.

This project will build on existing incremental speech processing techniques in the context of fast-paced interactive dialogue games, where speed is crucial to the system’s success. The work will focus specifically on the development of improved incremental language understanding techniques and dialogue policies that can enable rapid and appropriate system responses. Some of the relevant types of responses that might be addressed include verbal and non-verbal back-channels, interruptions, clarifications, collaborative completions, other grounding responses, or simply moving on to the next topic.

Job Description
The successful intern will learn about current techniques for incremental understanding and confidence modeling,
explore the application of these techniques in systems that play interactive dialogue games, define appropriate evaluation criteria, and perform an evaluation. This work will be a collaborative effort with several members of the ICT natural language dialogue group.

Preferred Skills

• Required: Master’s or PhD student working on dialogue systems or natural language processing
• Required: Sufficient programming ability to be comfortable with medium-sized programming projects
• Desired: Java programming experience
• Desired: Some familiarity with dialogue systems or natural language dialogue
• Desired: Familiarity and interest in incremental language processing techniques
• Desired: Interest in dialogue system policies

Project Name
Evaluation of Dialogue Policies with Virtual Humans

Project Description
A dialogue policy decides on which actions a dialogue system should perform given some dialogue context. Dialogue policies can be either hand-crafted or learned from data and simulations, e.g., using reinforcement learning, supervised learning, etc. The goal of this project is to (1) design and implement a new framework or improve an existing one for connecting dialogue policies to ICT’s virtual humans architecture, (2) evaluate such policies with human users, and (3) analyze the results of tests with human users. The dialogue policies will be provided to the intern by ICT researchers.

Job Description
The intern will work with the ICT natural language group and researchers/programmers from other ICT groups to help build a new framework or improve an existing one for connecting dialogue policies to ICT’s virtual humans architecture. This involves understanding the different dialogue system components and ICT’s virtual humans architecture. The intern will also evaluate existing dialogue policies with human users and statistically analyze these human-user interactions.

Preferred Skills

• Familiarity with dialogue systems and natural language dialogue
• Familiarity with statistical analysis of data
• Some knowledge of scientific experimental method (e.g., how to design an experiment)
• Good programming skills (preferably C++, Java, or Python)

Project Name
Virtual Humans Natural Language Dialogue

Project Description
ICT is developing artificial intelligence and natural language processing technology to allow virtual humans to engage in spoken and face to face interactions with people for a variety of purposes, including training of conversational tasks with virtual role-players. Current research areas include, Embodied, socio-cultural & affective dialogue, strategic dialogue, dialogue architectures, computational theories of dialogue genres, evaluation of dialogue systems, and dialogue authoring.

Job Description
The student intern will work with the Natural language research group (including Professors, other professional researchers, and students) to advance one or more of the research areas described above. If the student has a particular goal or related work at their home institution they should briefly describe this in the application letter.

Preferred Skills

• Some familiarity with dialogue systems or natural language dialogue
• Either programming ability or experience with statistical methods and data analysis
• Ability to work independently as well as in a collaborative environment

Project Name
High Speed Camera Project

Project Description
The goal is to dramatically shorten the Light Stage scanning time from currently 3s~20s to only 1/10s or even less by adopting 300fps+ high-speed image sensor with global shutter. The HSC will facilitate Light Stage scanning process by providing remote live-view, integrated control, hands-free data transfer and focus confirmation and image quality preview. It will also significantly reduce the bandwidth usage (for Light Stage and Headcam) by adding built-in pre-process algorithm on camera. Data volume generated by the latest high-speed image sensor is overflowing the capability of currently computer communication protocol such as USB2.0, Firewire 800, CameraLink or USB 3.0.

Job Description
Intern will initially work with another researcher on the image based lighting, high-speed image capture projects and on possibly automating the Light Stage facial scanning pipeline. Will participate in prototype development that involves electronics, firmware and software design in the field of digital photography and computer graphics.

Candidates with background of computer graphics, machine vision and software engineering are desired. Understanding of electrical engineering is a plus.

Preferred Skills

• programming with C/C++, Pyhton and Matlab
• knowledge of computer communication protocol such as USB2.0/3.0, HDMI, Camera-Link, etc
• experience in hardware programming(Verilog/VHDL, micro-controller) is a plus

Project Name
Virtual Humans Natural Language

Project Description
This project will involve taking one or more state of the art models of pheneomena in linguistic pragmatics, (e.g., speech act update semantics, reference resolution, presupposition, implicature, relevance) that the student has been working on, and working with ICT researchers to implement the approach within an advanced dialogue system for multi-modal dialogue.

Job Description
he intern will work within a collaborative group of faculty, PhD researchers and students on advancing the capabilities of virtual human dialogue by importing new approaches to pragmatic issues to advance the capabilities of a virtual human conversation system. The work will include adapting the theory to make use of available resources within the system, helping implement the theory, and evaluating performance. Some work on data collection, either for training or evaluation may also be needed.

Preferred Skills

• Experience working in one or more aspects of pragmatics of dialogue.
• Some familiarity with programming and/or computational models.
• Ability to work independently as well as in a collaborative environment.

Project Name
Full Body Performance Acquisition Project

Project Description
Over the last twenty years, direct interaction between Holocaust survivors and the museum public and school students has significantly furthered knowledge and understanding about the Holocaust. One person’s direct conversation with another is a particularly intimate and powerful method for educating and inspiring. Given that in a few years Holocaust survivors will no longer be with us to share their experiences first hand, finding a way to capture, preserve and present their testimonies in an authentic and interactive manner is paramount. The New Dimensions in Holocaust Testimony project aims to provide a way to capture the dialogue between survivors and visitors by recording living survivors using the latest 3D imaging technology. The project will display these recordings in a 3D interactive manner, utilizing natural language understanding, to create an immersive environment as realistic and natural as possible. As in the museum or classroom, viewers will listen to a brief summary talk followed by interactive question and answering. Consequently, the conversations that are taking place now between survivors and students can be reproduced in museums, over the web and in classrooms for many years to come.

Job Description
This particular internship is focused on using the latest computer vision techniques to accelerate the video capture and processing of a single survivor interview. Data will be captured with large number of video cameras for several hours. The intern will assist in developing new stereo and optical flow algorithms to process this data and allow for real-time view interpolation.

Preferred Skills

• C/C++,
• GPU programming
• Computer vision techniques in particular: multi-camera stereo, optical flow, and view interpolation Video compression (MPEG)
• Experience with OpenGL

Project Name
Stress Resilience in Virtual Environments (STRIVE)

Project Description
STRIVE involves basic research into physiological, psychological and biological mechanisms in stress reactivity and resilience. This novel research is made possible by the advanced series of immersive episodes developed under the direction of Dr. Skip Rizzo. These cinematically advanced episodes are designed to elicit specific emotions through narrative, visual and audio stimuli. Each episode also includes a virtual mentor who provides a basic understanding of the physiology and psychology of resilience. Evaluation of responses to these scenarios includes EEG, EKG, GSR and respiration. Baseline effects of stress are determined by advanced analysis of a series of over 20 biomarkers known to be affected by stress. This network of biomarkers, referred to as allostatic load, will be evaluated to determine if there are profiles associated with effective and dysfunctional acute stress responses.

Job Description
This Research Assistant will assist the Project PI, Dr. Galen Buckwalter, in the analysis of a comprehensive dataset collected from Marines at Camp Pendleton. The RA will be responsible for basic data management, data cleaning, and preliminary statistical analyses. The RA will have the opportunity to participate in more in-depth data analyses should that be an area of interest, including Independent Component Analysis of EEG data. The determined RA may have the opportunity to participate in manuscript preparation as well.

Preferred Skills

• Commitment and energy
• Computer skills
• Understanding of basic statistics

Project Name
Natural Language Annotation/Transcription

Project Description
The Natural Language Dialogue team collects linguistic data for use in developing, training and extending coverage of our conversational dialogue systems. The annotation/transcription project consists of transcribing speech and annotating conversation transcripts, collected through role-plays and interaction with existing systems.

Job Description
Transcription of English speech and annotation of conversation transcripts, using semantic representations that have been developed specifically for our implemented systems. The job is suitable primarily for undergraduate students. Interns who reside locally in the Los Angeles area may be able to continue working at ICT after the summer.

Preferred Skills

• Very good spoken and written English (native or near-native competence preferred).
• Some background in Linguistics or a related field.
• General feel for language and working with linguistic material.

Project Name
Spoken Language Understanding for Conversational Dialogue Systems

Project Description
Spoken dialogue systems typically include separate modules for automatic speech recognition (ASR) and natural language understanding (NLU). This project aims to improve overall understanding of speech input by investigating integration of ASR and NLU in ways that go beyond a strict pipeline, abandoning the assumption that speech-to-words and words-to-meaning mappings are performed independently. Ideas that may be explored include: NLU-based loss functions for discriminative language modeling, phonetic information for NLU, and syntactic, semantic and pragmatic information for ASR. This research will be performed using existing annotated data from ICT’s virtual human characters.

Job Description
The intern will use data-driven techniques for speech and language processing to perform experiments using audio and transcript data from ICT’s virtual human systems.

Preferred Skills

• Required: familiarity with natural language processing or speech recognition research.
• Required: strong programming skills.
• Desired: Standard Unix text processing utilities, Perl, Python or shell scripts.

Project Name
Integrated Virtual Humans

Project Description
The Integrated Virtual Humans project (IVH) seeks to create a wide range of virtual human systems by combining the various research efforts within USC and ICT into a general Virtual Human Architecture. These virtual humans range from relatively simple, statistics based question / answer characters to advanced, cognitive agents that are able to reason about themselves and the world they inhabit. Our virtual humans can engage with real humans and each other both verbally and nonverbally, i.e., they are able to hear you, see you, use body language, talk to you, and think about whether or not they like you. The Virtual Humans research at ICT is widely considered one of the most advanced in its field and brings together a variety of research areas, including natural language processing, nonverbal behavior, vision perception and understanding, task modeling, emotion modeling, information retrieval, knowledge representation, and speech recognition.

Job Description
Some of the challenges when developing virtual humans are the complexity of the system and the amount of specialized knowledge one needs in order to create new agents. Tools that support the authoring and debugging of agents are therefore essential, but in no way trivial to develop. For instance, how would you visualize the agent’s state of mind, taking into consideration that this involves its emotions, its views of others, its desires and intentions, its beliefs about possible actions and their consequences, and the current state of the conversation?

Considering this challenge, the tasks outlined for the summer internship are as follows:

• Become familiar with the general Virtual Humans Architecture and interact with several virtual humans;
• Create a new virtual human within a small domain, using existing authoring and debug tools;
• Give feedback on these tools;
• Implement this feedback in existing tools and/or create new tools.

Working within IVH requires a solid understanding of general software engineering principles and distributed architectures. The work touches on a variety of Computer Science areas, including Artificial Intelligence and Human-Computer Interaction. Given the scope of the Virtual Humans Architecture, the ability to quickly learn how to use existing components and develop new ones is essential.

Preferred Skills

• Fluent in C++, C# or Java
• Fluent in one or more scripting languages, like Python, TCL, LUA, or PHP
• Excellent general computer skills
• Background in Artificial Intelligence a plus

Project Name
Examining the Interplay between Emotions and Decision Making in Social Networks

Project Description
USC’s Emotion Project examines the role of emotion in human cognition and human-computer interaction and applies computational techniques the modeling, recognition, and synthesis of emotional phenomena. The goal of this project is to extend contemporary attempts for detecting linguistic features related to emotions and sentiment by coupling this research with a distinct theoretical emphasis on the underlying cognitive factors that influence emotional rhetoric and, perhaps more importantly, how these factors unfold over time in (sometimes) predictable patterns. Importantly, as social media is utilized in a wide range of linguistic and cultural contexts, the emphasis of the research will be on data-driven methods that can be trained on text corpora from any culture and language. Our approach will make it possible to identify and track social media content that reveals specific traits in a specific population, among the vast and seemingly insurmountable volume of user-generated content published daily. This project will both help us better understand the interplay between affect and decision making in a cultural context, and it will feed into the development of advanced virtual humans that are able to simulate realistic emotional processing and potentially influence the outcome of negotiations with human users.

Job Description
The Emotion Project is seeking a research intern that will collaborate with a multi-university, interdisciplinary team on the development of a computational model of emotion. The intern is expected to play a central role in researching, designing and conducting psychological experiments to investigate the role of emotions in decision making. This requires extensive background in designing, preforming and analyzing behavioral experiments. The intern also has the opportunity to collaborate with a group of cognitive scientists and computer scientists to facilitate integration of a computational model of decision making into the Virtual Human Toolkit. The ideal candidate will be a PhD student in psychology or cognitive science.

Preferred Skills

• Social network analysis
• Natural language processing
• Social psychological theories of emotion and interpersonal interaction

Project Name
Integration of Analogical Reasoning into the Virtual Human Architecture

Project Description
An important goal for intelligent virtual agents is the ability to interact naturally with human users. Humans rely extensively on analogical reasoning and remembered experiences (personal or otherwise) when interacting with other humans. When a conversation participant describes an experience, other human participants are able to relate that situation to analogous experiences. Those experiences inform interpretation and response, impacting the conversation itself, as well as the mental, emotional and relational outcomes from it. The goal of this project would be to develop a method and technology for analogical reasoning through Virtual Humans using the Structure-Mapping Engine (SME) (Falkenhainer, Forbus, & Gentner, 1989), a cognitively plausible model of human similarity and analogical reasoning. This project will facilitate integration of SME in to the Virtual Human Toolkit for utilizing analogical processing in a wide range of applications related to Virtual Human research.

Job Description
The Emotion Project is seeking a research intern that will collaborate with a multi-university, interdisciplinary team. The intern is expected to play a central role in integrating the Structure-Mapping Engine in to the Virtual Human Toolkit. This requires extensive engineering work for building connections between SME and Soar cognitive architecture. The intern will have the opportunity to explore a variety of applications in which this integration can be used in, such as analogical story-telling through virtual humans. The ideal candidate will be a graduate student in computer science.

Preferred Skills

• Cognitive modeling
• Knowledge about Cognitive architectures, particularly SOAR
• Good programming skills in Lisp

Project Name
Multimodal Fusion and Behavior Understanding

Project Description
During face-to-face conversations, people naturally integrate speech and gestures to understand the message communicated by other participants. This last decade has seen a boom in multimodal research. While computers are now able to recognize speech and vision individually (up to a certain point!), the focus is now on the integration of these different sources of information: language, vocalization, visual behaviors. The goal of this project is to build new computational model of multimodal integration and behavior understanding. One important aspect of this project is the study of differences due to cultures, personalities and gender.

Job Description
The internship will be performed under the supervision of Prof. Louis-Philippe Morency, director of ICT MultiComp Lab. Using some of the existing public datasets (AMI, SEMAINE) as well as some of ICT’s datasets, the goal of this internship is to create new multimodal probabilistic models able recognize high-level behaviors such as emotion, dominance and turn-taking. To achieve this goal, the intern will need to learn about the current state-of-the-art in multimodal integration (e.g., Latent Dynamic CRF) as well as the literature background from social sciences about human behaviors understanding (e.g., David McNeill). The emphasis of this internship will be on integrating language and nonverbal behaviors.

Preferred Skills

• Ph.D. candidate
• Strong knowledge of machine learning and human communication
• Programming experience in C++ and Matlab
• Knowledge of human nonverbal communication

Project Name
The Sigma Cognitive Architecture

Project Description
This project is developing a new cognitive architecture; i.e., a computational hypothesis about the core mechanisms and their interactions that underly intelligent behavior, whether natural or artificial. This is in support of work on virtual humans at ICT – as well as eventually of intelligent robots/agents – and is based on the elegant but powerful formalism of graphical models (factor graphs). Work is in progress on memory, decision making, problem solving, learning, perception, localization and mapping, mental imagery, natural language, speech, and theory of mind.

Job Description
Looking for a student interested in developing, applying (to sample tasks), analyzing and/or evaluating new intelligent capabilities in an architectural framework.

Preferred Skills

• Programming (Lisp preferred, but can be learned once arrive)
• Graphical models (experience preferred, but ability to learn quickly is essential)
• Cognitive architectures (experience preferred, but interest is essential)