Difference between revisions of "Projects"
From Immersive Visualization Lab Wiki
(→XAI (Jurgen Schulze, 2017-2021)) |
(→XAI (Jurgen Schulze, 2017-2021)) |
||
| Line 16: | Line 16: | ||
<td>The effectiveness of AI systems is limited by the machine’s current inability to explain their decisions and actions to human users. The Department of Defense (DoD) is facing challenges that demand more intelligent, autonomous, and symbiotic systems. The Explainable AI (XAI) program aims to create a suite of machine learning techniques that produce more explainable models, while maintaining a high level of learning performance (prediction accuracy); and enable human users to understand, appropriately trust, and effectively manage the emerging generation of artificially intelligent partners. This project is a collaboration with the [https://www.sri.com/ Stanford Research Institute (SRI) Princeton, NJ]. Our part is the development of the web interface, as well as in-person user studies, for which we recruited over 100 subjects so far.<br> | <td>The effectiveness of AI systems is limited by the machine’s current inability to explain their decisions and actions to human users. The Department of Defense (DoD) is facing challenges that demand more intelligent, autonomous, and symbiotic systems. The Explainable AI (XAI) program aims to create a suite of machine learning techniques that produce more explainable models, while maintaining a high level of learning performance (prediction accuracy); and enable human users to understand, appropriately trust, and effectively manage the emerging generation of artificially intelligent partners. This project is a collaboration with the [https://www.sri.com/ Stanford Research Institute (SRI) Princeton, NJ]. Our part is the development of the web interface, as well as in-person user studies, for which we recruited over 100 subjects so far.<br> | ||
Publications: | Publications: | ||
| − | * Alipour, K., Schulze, J.P., Yao, Y., Ziskind, A., Burachas, G., "A Study on Multimodal and Interactive Explanations for Visual Question Answering", In Proceedings of the Workshop on Artificial Intelligence Safety (SafeAI 2020), New York, NY, Feb 7, 2020 | + | * Alipour, K., Schulze, J.P., Yao, Y., Ziskind, A., Burachas, G., [http://web.eng.ucsd.edu/~jschulze/publications/Alipour2020a.pdf "A Study on Multimodal and Interactive Explanations for Visual Question Answering"], In Proceedings of the Workshop on Artificial Intelligence Safety (SafeAI 2020), New York, NY, Feb 7, 2020 |
| − | * Alipour, K., Ray, A., Lin, X, Schulze, J.P., Yao, Y., Burachas, G.T., "The Impact of Explanations on AI Competency Prediction in VQA", In Proceedings of the IEEE conference on Humanized Computing and Communication with Artificial Intelligence (HCCAI 2020), Sept 22, 2020 | + | * Alipour, K., Ray, A., Lin, X, Schulze, J.P., Yao, Y., Burachas, G.T., [http://web.eng.ucsd.edu/~jschulze/publications/Alipour2020b.pdf "The Impact of Explanations on AI Competency Prediction in VQA"], In Proceedings of the IEEE conference on Humanized Computing and Communication with Artificial Intelligence (HCCAI 2020), Sept 22, 2020 |
</td> | </td> | ||
</tr> | </tr> | ||
Revision as of 00:54, 15 December 2020
Contents |
Current Projects
Helmsley (Larry Smarr, Jurgen Schulze, 2018-2021)
 |
UCOP funded Catalyst project on cyber-archaeology. “UC San Diego’s Thomas E. Levy Among Recipients of President’s Research Catalyst Award.” UC San Diego News Center, December 9. http://ucs-dnews.ucsd.edu/pressrelease/uc_san_diegos_thomas_e._levy_among_recipi-ents_of_presidents_research_cataly. |
XAI (Jurgen Schulze, 2017-2021)
 |
The effectiveness of AI systems is limited by the machine’s current inability to explain their decisions and actions to human users. The Department of Defense (DoD) is facing challenges that demand more intelligent, autonomous, and symbiotic systems. The Explainable AI (XAI) program aims to create a suite of machine learning techniques that produce more explainable models, while maintaining a high level of learning performance (prediction accuracy); and enable human users to understand, appropriately trust, and effectively manage the emerging generation of artificially intelligent partners. This project is a collaboration with the Stanford Research Institute (SRI) Princeton, NJ. Our part is the development of the web interface, as well as in-person user studies, for which we recruited over 100 subjects so far. Publications:
|
DataCube (Larry Smarr, Jurgen Schulze, 2018-2021)
|
UCOP funded Catalyst project on cyber-archaeology. “UC San Diego’s Thomas E. Levy Among Recipients of President’s Research Catalyst Award.” UC San Diego News Center, December 9. http://ucs-dnews.ucsd.edu/pressrelease/uc_san_diegos_thomas_e._levy_among_recipi-ents_of_presidents_research_cataly. |
Catalyst (Tom Levy, Jurgen Schulze, 2017-2019)
|
UCOP funded Catalyst project on cyber-archaeology. “UC San Diego’s Thomas E. Levy Among Recipients of President’s Research Catalyst Award.” UC San Diego News Center, December 9. http://ucs-dnews.ucsd.edu/pressrelease/uc_san_diegos_thomas_e._levy_among_recipi-ents_of_presidents_research_cataly. |
CalVR (Andrew Prudhomme, Philip Weber, Jurgen Schulze, since 2010)
 |
CalVR is our virtual reality middleware (a.k.a. VR engine), which we have been developing for our graphics clusters. It runs on anything from a laptop to a large multi-node CAVE, and builds under Linux, Windows and MacOS. More information about how to obtain the code and build it can be found on our main CalVR page. We also wrote a paper on CalVR, and gave a presentation on it. |
