output
20022017
most citedCaffe: Convolutional Architecture for Fast Feature Embedding

4.3k citations

Showing 2017Show all

15 papers · 1 filter

cs.CV201741 cited

Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions

Bichen Wu, Alvin Wan, Xiangyu Yue +6

Neural networks rely on convolutions to aggregate spatial information. However, spatial convolutions are expensive in terms of model size and computation, both of which grow quadra…

cs.DS20175 cited

Heavy Hitters and the Structure of Local Privacy

Mark Bun, Jelani Nelson, Uri Stemmer

We present a new locally differentially private algorithm for the heavy hitters problem which achieves optimal worst-case error as a function of all standardly considered parameter…

astro-ph.CO201710 cited

Subpercent Accurate Fitting of Modified Gravity Growth

Mikhail Denissenya, Eric V. Linder

Adding to our previous method for dealing with gravitational modifications at redshift through a single parameter, we investigate treatment of lower redshift modificati…

astro-ph.IM20172 cited

All-sky Radio SETI

Michael Garrett, Andrew Siemion, Wim van Cappellen

Over the last decade, Aperture Arrays (AA) have successfully replaced parabolic dishes as the technology of choice at low radio frequencies - good examples are the MWA, LWA and LOF…

cs.CV201745 cited

ImageNet Training in Minutes

Yang You, Zhao Zhang, Cho-Jui Hsieh +2

Finishing 90-epoch ImageNet-1k training with ResNet-50 on a NVIDIA M40 GPU takes 14 days. This training requires 10^18 single precision operations in total. On the other hand, the…

stat.ML2017

Low Permutation-rank Matrices: Structural Properties and Noisy Completion

Nihar B. Shah, Sivaraman Balakrishnan, Martin J. Wainwright

We consider the problem of noisy matrix completion, in which the goal is to reconstruct a structured matrix whose entries are partially observed in noise. Standard approaches to th…