Artificial intelligence adversarial vulnerability audit tool
US Patent: 10839268
US Patent: 10839268
US Patent App: 16710640
US Patent: 10846407
Affective Computing (CSCI 534) class project. System-built on top of FAIR negotiation bot
Class project completed for UCSC Mechatronics course in which we had to design and build a robot which completed the SlugWars challenge.
Published in IEEE Virtual Reality 2017, 2017
This paper documents the work surrounding early data visualization and mission planning in VR for NASA JPL
Recommended citation: V. Ardulov and O. Pariser, "Immersive data interaction for planetary and earth sciences," 2017 IEEE Virtual Reality (VR), Los Angeles, CA, 2017, pp. 263-264. https://ieeexplore.ieee.org/abstract/document/7892277
Published in arXiv, 2018
This paper introduces a novel computational method for tracking the verbal and conversational productivity of a child during forensic interviewing for legal procedings.
Recommended citation: V. Ardulov, M. Kumar, S. Williams, T. Lyon, and S. Narayanan. 2018. Measuring Conversational Productivity in Child Forensic Interviews. ArXiv e-prints (June 2018). arXiv:cs.CL/1806.0335 https://arxiv.org/abs/1806.03357
Published in ACM International Conference on Multimodal Interactions, 2018
This paper looks at multimodal (language, affect, and speech) models of child forensic interviewing. In particular we explore Linear Mixture Models (LMMs) and Dynamic Mode Decomposition with Control (DMDc)
Recommended citation: Victor Ardulov, Madelyn Mendlen, Manoj Kumar, Neha Anand, Shanna Williams, Thomas Lyon, and Shrikanth Narayanan. 2018. Multimodal Interaction Modeling of Child Forensic Interviewing. In Proceedings of the 20th ACM International Conference on Multimodal Interaction (ICMI 18). ACM, New York, NY, USA, 179-185. DOI: https://doi.org/10.1145/3242969.3243006 https://dl.acm.org/citation.cfm?doid=3242969.3243006
Published in InterSpeech 2019, 2019
This works explores the use of frame psychotherapy as story-telling process, and looks at the estimation of therapist-client alliance as a product of the personae that each occupies during the therapy seesion.
Recommended citation: Martinez, V.R., Flemotomos, N., Ardulov, V., Somandepalli, K., Goldberg, S.B., Imel, Z.E., Atkins, D.C., Narayanan, S. (2019) Identifying Therapist and Client Personae for Therapeutic Alliance Estimation. Proc. Interspeech 2019, 1901-1905, DOI: 10.21437/Interspeech.2019-2829. https://www.isca-speech.org/archive/Interspeech_2019/pdfs/2829.pdf
Published in ICASSP 2020, 2020
Building on existing approaches, we analyze how the vocabulary and emotional content and language use of a child indicates truthfulness pointing to specific psycho-linguistic features that models correlate with truthfulness.
Recommended citation: V. Ardulov, Z. Durante, S. Williams, T. Lyon and S. Narayanan, "Identifying Truthful Language in Child Interviews," ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain, 2020, pp. 8074-8078, doi: 10.1109/ICASSP40776.2020.9053386. https://ieeexplore.ieee.org/abstract/document/9053386
Published in arXiv, 2020
We analyze the use of MI codes as a intermediate feature to improve the automatic prediction of CBT adherence via CTRS score
Recommended citation: Chen, Z., Flemotomos, N., Ardulov, V., Creed, T. A., Imel, Z. E., Atkins, D. C., & Narayanan, S. (2020). Feature Fusion Strategies for End-to-End Evaluation of Cognitive Behavior Therapy Sessions. arXiv preprint arXiv:2005.07809. https://arxiv.org/pdf/2005.07809
Published in ICMI 2020, 2020
This work evaluates the multimodal coordination of children across the still face interaction across ages
Recommended citation: Klein, L., Ardulov, V., Hu, Y., Soleymani, M., Gharib, A., Thompson, B., ... & Matarić, M. J. (2020, October). Incorporating Measures of Intermodal Coordination in Automated Analysis of Infant-Mother Interaction. In Proceedings of the 2020 International Conference on Multimodal Interaction (pp. 287-295). https://dl.acm.org/doi/abs/10.1145/3382507.3418870
Published in arXiv, 2021
An system overview of an end-to-end automated psychotherapy evaluation pipeline developed by a large multi-year collaboration
Recommended citation: Flemotomos, N., Martinez, V. R., Chen, Z., Singla, K., Ardulov, V., Peri, R., ... & Narayanan, S. (2021). "Am I A good therapist?" automated evaluation of psychotherapy skills using speech and language technologies. CoRR, abs/2102.11265. https://arxiv.org/abs/2102.11265
Published in Scientific Reports (Nature), 2021
Through the paradigm of discerning ADHD and ASD we show the advantages of considering diagnostic classification as a decision-making process rather than a traditional machine learning problem
Recommended citation: Ardulov, V., Martinez, V. R., Somandepalli, K., Zheng, S., Salzman, E., Lord, C., ... & Narayanan, S. (2021). Robust diagnostic classification via Q-learning. Scientific reports, 11(1), 1-9. https://www.nature.com/articles/s41598-021-90000-4
Published in Computer Speech and Language, 2021
Through the utilization of Granger Causal Analysis on speech signals between a child and an adult interviewer, we are able to use this to discern when a child is telling the truth about a non-disclosure
Recommended citation: Zane Durante, Victor Ardulov, Manoj Kumar, Jennifer Gongola, Thomas Lyon, Shrikanth Narayanan, Causal indicators for assessing the truthfulness of child speech in forensic interviews, Computer Speech & Language, Volume 71, 2022, 101263, ISSN 0885-2308, https://doi.org/10.1016/j.csl.2021.101263. https://www.sciencedirect.com/science/article/pii/S0885230821000693
Published in International Conference on Multimodal Interaction 2021, 2021
Using a linear dynamical system model to identify behaviors correlated with social coordination between infants and their mothers
Recommended citation: Klein, L., Ardulov, V., Gharib, A., Thompson, B., Levitt, P., & Matarić, M. (2021, October). Dynamic Mode Decomposition with Control as a Model of Multimodal Behavioral Coordination. In Proceedings of the 2021 International Conference on Multimodal Interaction (pp. 25-33). https://dl.acm.org/doi/abs/10.1145/3462244.3479916
Published:
Slides from a fun little talk I gave in front of Seminar course in Computer Science. I am trying to extract wisdom from AI to help me with my not-so-artificial intelligence.
Published:
Slides from the presentation on my paper at ICMI
Published:
Slides from a talk given to Olga White’s BECA class at SFSU
Published:
A recording of the talk I submitted along side our paper to ICASSP 2020 when it was converted into a virtual conference:
Published:
Slides from a talk given to Dr. Ashley Larsson’s BECA class at SFSU
Laboratory, University of Southern California, Computer Science, 2019
Ran weekly laboratory sections geared towards introducing computer science students to the fundamentals of robotics. Coursework was designed for the application of many concepts from computer science in new application domains. Duties included preparing weekly lab environments, helping students work through problems as they arose, graded and evaluated labs and homework.
Undergraduate course, University of Southern California, Computer Science, 2020
Class covered fundamentals of AI starting from deterministic search-algorithms going to heuristic functions, statistical learning methods, eventually building up towards the basics of neural based approaches. Duties encompassed preparing programming assignments, writing solutions to homework and exams, monitoring student question and discussion board, and leading weekly discussion sections providing additional support and exploring tangential topics beyond those covered in class. This was a particularly challenging teaching experience as it occured during the early stages of COVID-19 and the instructor was new to the course, so we had to be creative, adaptive, and compassionate as we navigated the course and topics.