Mackenzie Leake

Mackenzie Leake

I am a PhD student in computer science at Stanford University advised by Maneesh Agrawala. In 2015 I received my BA in computational science and studio art at Scripps College. In the past I've worked on research projects at Columbia University, IBM Research Almaden, and Harvey Mudd College. My work is currently funded by the Brown Institute for Media Innovation and Adobe Research.

Contact Information:

first initial + lastname @cs.stanford.edu

Recent Projects

Roughcut
A tool that uses the structure of dialogue-driven scenes and cinematographic principles to edit video sequences automatically. The input is a script and multiple video takes, capturing different camera framings or performances of the scene. Our system automatically selects a clip from one of the takes, for each line of dialogue, based on film-editing idioms.
(SIGGRAPH paper | video | website)
CSEd
Recommendations for designing online communities for high school computer science teachers. Based upon interviews with high school computer science teachers, we propose strategies for improving online resources for computer science teachers.
(CSCW paper | SIGCSE paper)
HTM
Analyzing the learning dynamics for hierarchical temporal memory (HTM), a brain-inspired machine-learning paradigm. We discuss how the selection of input parameters influences the learning behavior of the spatial pooling phase of the HTM algorithm.
(AAAI paper | ICANN paper)

Publications

  • Leake, M., Davis, A., Truong, A., & Agrawala, M. (2017). Computational Video Editing for Dialogue-Driven Scenes. ACM Transactions on Graphics (SIGGRAPH'17), 36(4). ACM, New York, NY, USA, 130:1-130:14.
  • Leake, M. & Lewis, C.M. (2017). Designing CS Resource Sharing Sites for All Teachers. Proceedings of the 48th Technical Symposium on Computer Science Education (SIGCSE'17). ACM, New York, NY, USA, 357-362.
  • Leake, M. & Lewis, C.M. (2016). Designing a New System for Sharing Computer Science Teaching Resources. Proceedings of the 19th ACM Conference on Computer Supported Cooperative Work & Social Computing Companion (CSCW'16). ACM, New York, NY, USA, 321-324.
  • Leake, M. (2015). Chart-based Strategies for Solving Bayesian Inference Problems. Journal of Computing Sciences in Colleges. 30(4), Consortium for Computing Sciences in Colleges, USA. 107-114.
  • Leake, M., Xia, L., Rocki, K., & Imaino, W. (2015). A Probabilistic View of the Spatial Pooler in Hierarchical Temporal Memory. International Journal of Computer, Electrical, Automation, Control and Information Engineering, 9(5), 1111-1118.
  • Leake, M., Xia, L., Rocki, K., & Imaino, W. (2015). Effect of Spatial Pooler Initialization on Column Activity in Hierarchical Temporal Memory. 29th AAAI Conference on Artificial Intelligence (AAAI'15), 4176-4177.

Awards

  • Adobe Research Fellowship, 2017
  • Brown Institute for Media Innovation Magic Grant, 2016
  • Stanford EDGE-STEM Fellowship, 2015
  • Stanford School of Engineering Fellowship, 2015
  • Goldwater Scholar, 2014
  • Amgen Scholar, 2013
  • James E. Scripps Scholarship, 2011-2015