Publications

@inproceedings{devault_method_2013,

title = {A method for the approximation of incremental understanding of explicit utterance meaning using predictive models in finite domains},

author = {David DeVault and David Traum},

url = {http://ict.usc.edu/pubs/A%20method%20for%20the%20approximation%20of%20incremental%20understanding%20of%20explicit%20utterance%20meaning%20using%20predictive%20models%20in%20%EF%AC%81nite%20domains.pdf},

year  = {2013},

date = {2013-06-01},

booktitle = {Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},

address = {Atlanta, GA},

abstract = {This paper explores the relationship between explicit and predictive models of incremental speech understanding in a dialogue system that supports a finite set of user utterance meanings. We present a method that enables the approximation of explicit understanding using information implicit in a predictive understanding model for the same domain. We show promising performance for this method in a corpus evaluation, and discuss its practical application and annotation costs in relation to some alternative approaches.⬚⬚⬚},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{morie_what_2013,

title = {What Can Your Avatar Tell the Doctor?},

author = {Jacquelyn Morie and Sin-Hwa Kang},

url = {http://ict.usc.edu/pubs/What%20Can%20Your%20Avatar%20Tell%20the%20Doctor.pdf},

year  = {2013},

date = {2013-06-01},

booktitle = {The CyberPsychology & CyberTherapy Conference},

address = {Brussels, Belgium},

abstract = {The goal of our study is to explore a functioning virtual self that can present salient information about our health. We investigate several types of connectivity between a user and his/her animated character representation (also referred to as an avatar). We aim to find various forms of avatar connection with its user to address the value of avatars in health care applications. We propose an additional variation on an existing categorization of avatar-self associations. We specifically examine the potential of the biologically instrumented end of the avatar connectivity continuum.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{song_action_2013,

title = {Action Recognition by Hierarchical Sequence Summarization},

author = {Yale Song and Louis-Philippe Morency and Randall Davis},

url = {http://ict.usc.edu/pubs/Action%20Recognition%20by%20Hierarchical%20Sequence%20Summarization.pdf},

year  = {2013},

date = {2013-06-01},

booktitle = {Computer Vision and Pattern Recognition (CVPR 2013)},

address = {Portland, OR},

abstract = {Recent progress has shown that learning from hierarchical feature representations leads to improvements in various computer vision tasks. Motivated by the observations that human activity data contains information at various temporal resolutions, we present a hierarchical sequence summarization approach for action recognition that learns multiple layers of discriminative feature representations. We build up a hierarchy dynamically and recursively by alternating sequence learning and sequence summarization. For sequence learning we use CRFs with latent variables to learn hidden spatio-temporal dynamics; for sequence summarization we group observations that have similar semantic meaning in the latent space. For each layer we learn an abstract feature representation through non-linear gate functions. This procedure is repeated to obtain a hierarchical sequence summary representation. We develop an efficient learning method to train our model and show that its complexity grows sublinearly with the size of the hierarchy. Experimental results show the effectiveness of our approach, achieving the best published results on the Ar-mGesture and Canal9 datasets.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{scherer_investigating_2013,

title = {Investigating the Speech Characteristics of Suicidal Adolescents},

author = {Stefan Scherer and John Pestian and Louis-Philippe Morency},

url = {http://ict.usc.edu/pubs/Investigating%20the%20Speech%20Characteristics%20of%20Suicidal%20Adolescents.pdf},

year  = {2013},

date = {2013-05-01},

booktitle = {International Conference on Acoustics, Speech, and Signal Processing (ICASSP)},

address = {Vancouver, Canada},

abstract = {Suicide is a very serious problem. In the United states it ranks as the second most frequent cause of death among teenagers between the ages of 12 and 17. In this work, we investigate speech characteristics of prosody as well as voice quality in a dyadic interview corpus with suicidal and non-suicidal adolescents. In these interviews the adolescents answer specifically designed questions. Based on this limited dataset, we reveal statistically significant differences in the speech patterns of suicidal adolescents within the investigated interview corpus. Further, we investigate the classification capabilities of machine learning approaches both on an utterance as well as an interview level. The work shows promising results in a speaker independent classification experiment based on only a dozen speech features. We believe that once the algorithms are refined and integrated with other methods, they may be of value to the clinician.⬚⬚⬚⬚⬚},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{weninger_speaker_2013,

title = {Speaker Trait Characterization in Web Videos: Uniting Speech, Language, and Facial Features},

author = {Felix Weninger and Claudia Wagner and Martin Wollmer and Bjorn Schuller and Louis-Philippe Morency},

url = {http://ict.usc.edu/pubs/Speaker%20Trait%20Characterization%20in%20Web%20Videos-%20Uniting%20Speech,%20Language,%20and%20Facial%20Features.pdf},

year  = {2013},

date = {2013-05-01},

booktitle = {International Conference on Acoustics, Speech, and Signal Processing (ICASSP)},

address = {Vancouver, Canada},

abstract = {We present a multi-modal approach to speaker characterization using acoustic, visual and linguistic features. Full realism is provided by evaluation on a database of real-life web videos and automatic feature extraction including face and eye detection, and automatic speech recognition. Different segmentations are evaluated for the audio and video streams, and the statistical relevance of Linguistic Inquiry and Word Count (LIWC) features is confirmed. In the result, late multimodal fusion delivers 73, 92 and 73 % average recall in binary age, gender and race classification on unseen test subjects, outperforming the best single modalities for age and race.⬚},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{wienberg_content-based_2013,

title = {Content-Based Similarity Measures of Weblog Authors},

author = {Christopher Wienberg and Melissa Roemmele and Andrew S. Gordon},

url = {http://ict.usc.edu/pubs/Content-Based%20Similarity%20Measures%20of%20Weblog%20Authors.PDF},

year  = {2013},

date = {2013-05-01},

booktitle = {ACM Web Science Conference},

address = {Paris, France},

abstract = {With recent research interest in the confounding roles of homophily and contagion in studies of social influence, there is a strong need for reliable content-based measures of the similarity between people. In this paper, we investigate the use of text similarity measures as a way of predicting the similarity of prolific weblog authors. We describe a novel method of collecting human judgments of overall similarity between two authors, as well as demographic, political, cultural, religious, values, hobbies/interests, personality, and writing style similarity. We then apply a range of automated textual similarity measures based on word frequency counts, and calculate their statistical correlation with human judgments. Our findings indicate that commonly used text similarity measures do not correlate well with human judgments of author similarity. However, various measures that pay special attention to personal pronouns and their context correlate significantly with different facets of similarity.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{graham_measurement-based_2013,

title = {Measurement-based Synthesis of Facial Microgeometry},

author = {Paul Graham and Borom Tunwattanapong and Jay Busch and Xueming Yu and Andrew Jones and Paul Debevec and Abhijeet Ghosh},

url = {http://ict.usc.edu/pubs/Measurement-based%20Synthesis%20of%20Facial%20Microgeometry.pdf},

year  = {2013},

date = {2013-05-01},

booktitle = {EUROGRAPHICS},

address = {Girona, Spain},

abstract = {Current scanning techniques record facial mesostructure with submillimeter precision showing pores, wrinkles, and creases. However, surface roughness continues to shape specular reflection at the level of microstructure: micron scale structures. Here, we present an approach to increase the resolution of mesostructure-level facial scans using microstructure examples digitized about the face. We digitize the skin patches using polarized gradient illumination and 10 mm resolution macro photography, and observe point-source reflectance measurements to characterize the specular reflectance lobe at this smaller scale. We then perform constrained texture synthesis to create appropriate surface microstructure per facial region, blending the regions to cover the whole entire face. We show that renderings of microstructure-augmented facial models preserve the original scanned mesostructure and exhibit surface reflections which are qualitatively more consistent with real photographs.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{levesque_sequential_2013,

title = {Sequential Emotion Recognition using Latent-Dynamic Conditional Neural Fields},

author = {Julien-Charles Levesque and Louis-Philippe Morency and Christian Gagne},

url = {http://ict.usc.edu/pubs/Sequential%20Emotion%20Recognition%20using%20Latent-Dynamic%20Conditional%20Neural%20Fields.pdf},

year  = {2013},

date = {2013-04-01},

booktitle = {IEEE Conference on Automatic Face and Gesture Recognition},

address = {Shanghai, China},

abstract = {A wide number of problems in face and gesture analysis involve the labeling of temporal sequences. In this paper, we introduce a discriminative model for such sequence labeling tasks. This model involves two layers of latent dynamics, each with their separate roles. The first layer, the neural network or gating layer, aims to extract non-linear relationships between input data and output labels. The second layer, the hidden-states layer, aims to model temporal substructure in the sequence by learning hidden-states and their transition dynamics. A new regularization term is proposed for the training of this model, encouraging diversity between hidden-states. We evaluate the performance of this model on an audiovisual dataset of emotion recognition and compare it against other popular methods for sequence labeling.⬚⬚},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{kang_users_2013,

title = {Users’ Socially Desirable Responding with Computer Interviewers},

author = {Sin-Hwa Kang and Jacquelyn Morie},

url = {http://ict.usc.edu/pubs/Users%E2%80%99%20Socially%20Desirable%20Responding%20with%20Computer%20Interviewers.pdf},

year  = {2013},

date = {2013-04-01},

booktitle = {Computer-Human Interaction Conference},

address = {Paris, France},

abstract = {In this paper, we explore how different types of computer interviewers and the amount of self- disclosure from the interviewers affect the quantity of socially desirable responses displayed by interviewees. Online surveys were delivered by computer interviewers. The computer interviewers included a text-based interface and an anthropomorphic character interface. The interviewers’ self-disclosure presented their social norm violations. Interview questions were in the form of socially desirable response items representing impression management in this study. The experimental design was a 2 (Interviewers’ type) x 2 (Interviewers’ self-disclosure versus no self-disclosure) factorial between-subjects experiment. The main dependent variable was whether users’ socially desirable responses were affected by the type of interviewer and that amount of self-disclosure provided by the interviewer. The preliminary findings present the potential for self-disclosing anthropomorphic characters to reduce the social desirability bias present in interviewees with high public self-consciousness in their self-disclosure.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{song_distribution-sensitive_2013,

title = {Distribution-Sensitive Learning for Imbalanced Datasets},

author = {Yale Song and Louis-Philippe Morency and Randall Davis},

url = {http://ict.usc.edu/pubs/Distribution-Sensitive%20Learning%20for%20Imbalanced%20Datasets.pdf},

year  = {2013},

date = {2013-04-01},

booktitle = {IEEE Conference on Automatic Face and Gesture Recognition},

address = {Shanghai, China},

abstract = {Many real-world face and gesture datasets are by nature imbalanced across classes. Conventional statistical learning models (e.g., SVM, HMM, CRF), however, are sensitive to imbalanced datasets. In this paper we show how an imbalanced dataset affects the performance of a standard learning algorithm, and propose a distribution-sensitive prior to deal with the imbalanced data problem. This prior analyzes the training dataset before learning a model, and puts more weight on the samples from underrepresented classes, allowing all samples in the dataset to have a balanced impact in the learning process. We report on two empirical studies regarding learning with imbalanced data, using two publicly available recent gesture datasets, the Microsoft Research Cambridge-12 (MSRC-12) and NATOPS aircraft handling signals datasets. Experimental results show that learning from balanced data is important, and that the distribution-sensitive prior improves performance with imbalanced datasets.⬚},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{gratch_felt_2013,

title = {Felt Emotion and Social Context Determine the Intensity of Smiles in a Competitive Video Game},

author = {Jonathan Gratch and Lin Cheng and Stacy C. Marsella and Jill Boberg},

url = {http://ict.usc.edu/pubs/Felt%20Emotion%20and%20Social%20Context%20Determine%20the%20Intensity%20of%20Smiles%20in%20a%20Competitive%20Video%20Game.pdf},

year  = {2013},

date = {2013-04-01},

booktitle = {10th IEEE International Conference on Automatic Face and Gesture Recognition},

address = {Shanghai, China},

abstract = {The present study uses automatic facial expression recognition software to examine the relationship between social context and emotional feelings on the expression of emotion, to test claims that facial expressions reflect social motives rather than felt emotion. To vary emotional feelings, participants en- gaged in a competitive video game. Deception was used to sys- tematically manipulate perceptions of winning or losing. To vary social context, participants played either with friends or strangers. The results support the hypothesis of Hess and col- leagues that smiling is determined by both factors. The results further highlight the value of automatic expression recognition technology for psychological research and provide constraints on inferring emotion from facial displays.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{scherer_automatic_2013,

title = {Automatic Behavior Descriptors for Psychological Disorder Analysis},

author = {Stefan Scherer and Giota Stratou and Marwa Mahmound and Jill Boberg and Jonathan Gratch and Albert Rizzo and Louis-Philippe Morency},

url = {http://ict.usc.edu/pubs/Automatic%20Behavior%20Descriptors%20for%20Psychological%20Disorder%20Analysis.pdf},

year  = {2013},

date = {2013-04-01},

booktitle = {IEEE Conference on Automatic Face and Gesture Recognition},

address = {Shanghai, China},

abstract = {We investigate the capabilities of automatic nonverbal behavior descriptors to identify indicators of psychological disorders such as depression, anxiety, and post-traumatic stress disorder. We seek to confirm and enrich present state of the art, predominantly based on qualitative manual annotations, with automatic quantitative behavior descriptors. In this paper, we propose four nonverbal behavior descriptors that can be automatically estimated from visual signals. We introduce a new dataset called the Distress Assessment Interview Corpus (DAIC) which includes 167 dyadic interactions between a confederate interviewer and a paid participant. Our evaluation on this dataset shows correlation of our automatic behavior descriptors with specific psychological disorders as well as a generic distress measure. Our analysis also includes a deeper study of selfadaptor and fidgeting behaviors based on detailed annotations of where these behaviors occur.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{xu_simple_2013,

title = {A Simple Method for High Quality Artist-Driven Lip Syncing},

author = {Yuyu Xu and Andrew W. Feng and Ari Shapiro},

url = {http://ict.usc.edu/pubs/A%20Simple%20Method%20for%20High%20Quality%20Artist-Driven%20Lip%20Syncing.pdf},

year  = {2013},

date = {2013-03-01},

booktitle = {ACM Symposium on Interactive 3D Graphics and Games},

address = {Orlando, FL},

abstract = {We demonstrate a real-time lip animation algorithmthat can be used to generate synchronized facial movements with audio generated from a text-to-speech engine or from recorded audio. Our method requires an animator to construct animations using a canonical set of visemes for all pairwise combinations of a reduced phoneme set (diphones). The diphone animations are then stitched together to construct the final animation. This method can be easily retargeted to new faces that use the same set of visemes. Thus, our method can be applied to any character that utilizes the same, small set of facial poses. In addition, our method is editable in that it allows an artist to directly and easily change specific parts of the lip animation algorithm as needed. Our method requires no learning, can work on multiple languages, and is easily replicated. We make available publicly animations for lip syncing English utterances. We present a study showing the subjective quality of our algorithm, and compare it to the results of a popular commercial software package.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{vasylevska_flexible_2013,

title = {Flexible Spaces: Dynamic Layout Generation for Infinite Walking in Virtual Environments},

author = {Khrystyna Vasylevska and Hannes Kaufmann and Mark Bolas and Evan Suma},

url = {http://ict.usc.edu/pubs/Flexible%20Spaces-%20Dynamic%20Layout%20Generation%20for%20Infinite%20Walking%20in%20Virtual%20Environments.pdf},

year  = {2013},

date = {2013-03-01},

booktitle = {IEEE Symposium on 3D User Interfaces},

address = {Orlando, FL},

abstract = {Redirected walking techniques enable natural locomotion through immersive virtual environments (VEs) that are larger than the real world workspace. Most existing techniques rely upon manipulating the mapping between physical and virtual motions while the layout of the environment remains constant. However, if the primary focus of the experience is on the virtual world’s content, rather than on its spatial layout, then the goal of redirected walking can be achieved through an entirely different strategy. In this paper, we introduce flexible spaces – a novel redirection technique that enables infinite real walking in virtual environments that do not require replication of real world layouts. Flexible spaces overcome the limitations and generalize the use of overlapping (impossible) spaces and change blindness by employing procedural layout generation. Our approach allows VE designers to focus on the content of the virtual world independent of the implementation details imposed by real walking, thereby making spatial manipulation techniques more practical for use in a variety of application domains.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{vasylevska_flexible_2013-1,

title = {Flexible Spaces: A Virtual Step Outside of Reality},

author = {Khrystyna Vasylevska and Hannes Kaufmann and Mark Bolas and Evan Suma},

url = {http://ict.usc.edu/pubs/Flexible%20Spaces-%20A%20Virtual%20Step%20Outside%20of%20Reality.pdf},

year  = {2013},

date = {2013-03-01},

booktitle = {IEEE Virtual Reality},

address = {Orlando, FL},

abstract = {In this paper we introduce the concept of flexible spaces –– a novel redirection technique that generalizes the use of overlapping (impossible) spaces and change blindness in an algorithm for dynamic layout generation. Flexible spaces is an impossible environment that violates the real world constancy in favor of providing the experience of seamless, unrestricted natural walking over a large-scale virtual environment (VE).},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}