Michael Gygli

icon indicating a website link https://twitter.com/GygliMichael

icon company Cerrion icon location Zurich, Switzerland icon location Computer Vision and AI Researcher, PhD, YC Founder, ExGoogler

Results 3 issues of Michael Gygli

Add SupSub and ModMod approximations for non submodular loss functions

More info in "Iyer, R., & Bilmes, J. (2012). Algorithms for approximate minimization of the difference between submodular functions, with applications. [arXiv Preprint](http://arxiv.org/abs/1207.0560)"

enhancement

H5PYDataset returns ndarray instead of list

2 comment

When using Variable-length data, the datatype returned by get_data is an ndarray (object type) of an ndarray containing the data itself, e.g. for images, it is an ndarray containing the...

Performance difference compared to the paper

6 comment

Hi Thanks for providing the tensorflow version of C3D. You mention Top-1 accuracy of 72.6% on validation, while they report 85.2% in the paper (on test). How come the performance...