Sahil SinglainHackerNoon.comA new kind of pooling layer for faster and sharper convergenceSome prior knowledge of conv nets is assumedOct 1, 20179Oct 1, 20179
Sahil SinglainTowards Data ScienceNormalized center loss for language modelingCaveat: Some knowledge of recurrent neural networks is assumed.Sep 22, 20171Sep 22, 20171
Sahil SinglainML ReviewExperiments with a New Loss Term Added to the Standard Cross entropyRecently I came across this idea of center loss described in this paper. You define the outputs from the second last layers of the neural…Sep 9, 20174Sep 9, 20174
Sahil SinglainTowards Data ScienceExperiments with a new kind of convolutionCaveat: Some prior knowledge of CNNs is assumed for this postSep 3, 20173Sep 3, 20173
Sahil SinglainTowards Data SciencePractical techniques for getting style transfer to workThose who are not familiar with style transfer can read this blog post summarizing style transfer.Aug 30, 20171Aug 30, 20171
Sahil SinglainML ReviewGetting Inception Architectures to Work with Style TransferStyle transfer typically requires neural networks with a well-developed hierarchy of features for calculating the loss. Thus for this…Aug 29, 20173Aug 29, 20173
Sahil SinglainTowards Data ScienceExperiments on different loss configurations for style transferAs I participated in Jeremy Howard’s excellent Cutting edge deep learning for coders, I was quite intrigued by style transfer. I wanted to…Aug 28, 20173Aug 28, 20173
Sahil SinglaIntroduction to Style transferCaveat: Some knowledge of convolutional neural networks is assumed.Aug 28, 2017Aug 28, 2017