# HG changeset patch # User Yoshua Bengio # Date 1275170617 14400 # Node ID d02d288257bf70369aff862c0f77759009e2e0fc # Parent d48a7777e4d8bec91efed62ea895d1bd726b74d0 redone bib style diff -r d48a7777e4d8 -r d02d288257bf writeup/nips2010_submission.tex --- a/writeup/nips2010_submission.tex Sat May 29 16:56:49 2010 -0400 +++ b/writeup/nips2010_submission.tex Sat May 29 18:03:37 2010 -0400 @@ -5,7 +5,7 @@ \usepackage{algorithm,algorithmic} \usepackage[utf8]{inputenc} \usepackage{graphicx,subfigure} -\usepackage{mlapa} +\usepackage[numbers]{natbib} \title{Generating and Exploiting Perturbed and Multi-Task Handwritten Training Data for Deep Architectures} \author{The IFT6266 Gang} @@ -45,7 +45,7 @@ \section{Introduction} Deep Learning has emerged as a promising new area of research in -statistical machine learning (see~\emcite{Bengio-2009} for a review). +statistical machine learning (see~\citet{Bengio-2009} for a review). Learning algorithms for deep architectures are centered on the learning of useful representations of data, which are better suited to the task at hand. This is in great part inspired by observations of the mammalian visual cortex, @@ -53,16 +53,16 @@ different representation. In fact, it was found recently that the features learnt in deep architectures resemble those observed in the first two of these stages (in areas V1 and V2 -of visual cortex)~\cite{HonglakL2008}. +of visual cortex)~\citep{HonglakL2008}. Processing images typically involves transforming the raw pixel data into new {\bf representations} that can be used for analysis or classification. For example, a principal component analysis representation linearly projects the input image into a lower-dimensional feature space. Why learn a representation? Current practice in the computer vision literature converts the raw pixels into a hand-crafted representation -(e.g.\ SIFT features~\cite{Lowe04}), but deep learning algorithms +e.g.\ SIFT features~\citep{Lowe04}, but deep learning algorithms tend to discover similar features in their first few -levels~\cite{HonglakL2008,ranzato-08,Koray-08,VincentPLarochelleH2008-very-small}. +levels~\citep{HonglakL2008,ranzato-08,Koray-08,VincentPLarochelleH2008-very-small}. Learning increases the ease and practicality of developing representations that are at once tailored to specific tasks, yet are able to borrow statistical strength @@ -77,9 +77,9 @@ It is also only recently that successful algorithms were proposed to overcome some of these difficulties. All are based on unsupervised learning, often in an greedy layer-wise ``unsupervised pre-training'' -stage~\cite{Bengio-2009}. One of these layer initialization techniques, +stage~\citep{Bengio-2009}. One of these layer initialization techniques, applied here, is the Denoising -Auto-Encoder~(DEA)~\cite{VincentPLarochelleH2008-very-small}, which +Auto-Encoder~(DEA)~\citep{VincentPLarochelleH2008-very-small}, which performed similarly or better than previously proposed Restricted Boltzmann Machines in terms of unsupervised extraction of a hierarchy of features useful for classification. The principle is that each layer starting from @@ -99,7 +99,7 @@ classifier better not only on similarly perturbed images but also on the {\em original clean examples}? \item Do deep architectures benefit more from such {\em out-of-distribution} -examples, i.e. do they benefit more from the self-taught learning~\cite{RainaR2007} framework? +examples, i.e. do they benefit more from the self-taught learning~\citep{RainaR2007} framework? \item Similarly, does the feature learning step in deep learning algorithms benefit more training with similar but different classes (i.e. a multi-task learning scenario) than a corresponding shallow and purely supervised architecture? @@ -110,7 +110,7 @@ This section describes the different transformations we used to stochastically transform source images in order to obtain data. More details can -be found in this technical report~\cite{ift6266-tr-anonymous}. +be found in this technical report~\citep{ift6266-tr-anonymous}. The code for these transformations (mostly python) is available at {\tt http://anonymous.url.net}. All the modules in the pipeline share a global control parameter ($0 \le complexity \le 1$) that allows one to modulate the @@ -130,7 +130,7 @@ maximum displacement for the lowest or highest pixel line is of $round(complexity \times 32)$.\\ {\bf Thickness}\\ -Morpholigical operators of dilation and erosion~\cite{Haralick87,Serra82} +Morpholigical operators of dilation and erosion~\citep{Haralick87,Serra82} are applied. The neighborhood of each pixel is multiplied element-wise with a {\em structuring element} matrix. The pixel value is replaced by the maximum or the minimum of the resulting @@ -156,7 +156,7 @@ \times complexity]$ and $c$ and $f$ $\sim U[-4 \times complexity, 4 \times complexity]$.\\ {\bf Local Elastic Deformations}\\ -This filter induces a "wiggly" effect in the image, following~\cite{SimardSP03}, +This filter induces a "wiggly" effect in the image, following~\citet{SimardSP03}, which provides more details. Two "displacements" fields are generated and applied, for horizontal and vertical displacements of pixels. @@ -171,7 +171,7 @@ {\bf Pinch}\\ This GIMP filter is named "Whirl and pinch", but whirl was set to 0. A pinch is ``similar to projecting the image onto an elastic -surface and pressing or pulling on the center of the surface''~\cite{GIMP-manual}. +surface and pressing or pulling on the center of the surface''~\citep{GIMP-manual}. For a square input image, think of drawing a circle of radius $r$ around a center point $C$. Any point (pixel) $P$ belonging to that disk (region inside circle) will have its value recalculated by taking @@ -198,7 +198,7 @@ closer to black. The corners of the occluder The rectangle corners are sampled so that larger complexity gives larger rectangles. The destination position in the occluded image are also sampled -according to a normal distribution (see more details in~\cite{ift6266-tr-anonymous}. +according to a normal distribution (see more details in~\citet{ift6266-tr-anonymous}). It has has a probability of not being applied at all of 60\%.\\ {\bf Pixel Permutation}\\ This filter permutes neighbouring pixels. It selects first @@ -212,7 +212,7 @@ noise $\sim Normal(0(\frac{complexity}{10})^2)$. It has has a probability of not being applied at all of 70\%.\\ {\bf Background Images}\\ -Following~\cite{Larochelle-jmlr-2009}, this transformation adds a random +Following~\citet{Larochelle-jmlr-2009}, this transformation adds a random background behind the letter. The background is chosen by first selecting, at random, an image from a set of images. Then a 32$\times$32 subregion of that image is chosen as the background image (by sampling position @@ -367,7 +367,7 @@ The stacked version is an adaptation to deep MLPs where you initialize each layer with a denoising auto-encoder starting from the bottom. During the initialization, which is usually called pre-training, the bottom layer is treated as if it were an isolated auto-encoder. The second and following layers receive the same treatment except that they take as input the encoded version of the data that has gone through the layers before it. -For additional details see \cite{vincent:icml08}. +For additional details see \citet{vincent:icml08}. \section{Experimental Results} @@ -379,8 +379,8 @@ service\footnote{http://mturk.com}. AMT users are paid small amounts of money to perform tasks for which human intelligence is required. Mechanical Turk has been used extensively in natural language -processing \cite{SnowEtAl2008} and vision -\cite{SorokinAndForsyth2008,whitehill09}. AMT users where presented +processing \citep{SnowEtAl2008} and vision +\citep{SorokinAndForsyth2008,whitehill09}. AMT users where presented with 10 character images and asked to type 10 corresponding ascii characters. Hence they were forced to make a hard choice among the 62 character classes. Three users classified each image, allowing @@ -408,10 +408,10 @@ MLP0 & 24.2\% $\pm$.15\% & 68.8\%$\pm$.33\% & 78.70\%$\pm$.14\% & 3.45\% $\pm$.15\% \\ \hline MLP1 & 23.0\% $\pm$.15\% & 41.8\%$\pm$.35\% & 90.4\%$\pm$.1\% & 3.85\% $\pm$.16\% \\ \hline MLP2 & 24.3\% $\pm$.15\% & 46.0\%$\pm$.35\% & 54.7\%$\pm$.17\% & 4.85\% $\pm$.18\% \\ \hline -\cite{Granger+al-2007} & & & & 4.95\% $\pm$.18\% \\ \hline -\cite{Cortes+al-2000} & & & & 3.71\% $\pm$.16\% \\ \hline -\cite{Oliveira+al-2002} & & & & 2.4\% $\pm$.13\% \\ \hline -\cite{Migram+al-2005} & & & & 2.1\% $\pm$.12\% \\ \hline +\citep{Granger+al-2007} & & & & 4.95\% $\pm$.18\% \\ \hline +\citep{Cortes+al-2000} & & & & 3.71\% $\pm$.16\% \\ \hline +\citep{Oliveira+al-2002} & & & & 2.4\% $\pm$.13\% \\ \hline +\citep{Migram+al-2005} & & & & 2.1\% $\pm$.12\% \\ \hline \end{tabular} \end{center} \end{table} @@ -427,7 +427,7 @@ from perturbed training data, even when testing on clean data, whereas the MLP trained on perturbed data performed worse on the clean digits and about the same on the clean characters. } -\label{tab:sda-vs-mlp-vs-humans} +\label{tab:perturbation-effect} \begin{center} \begin{tabular}{|l|r|r|r|r|} \hline & NIST test & NISTP test & P07 test & NIST test digits \\ \hline @@ -490,7 +490,8 @@ \section{Conclusions} \bibliography{strings,ml,aigaion,specials} -\bibliographystyle{mlapa} +%\bibliographystyle{plainnat} +\bibliographystyle{unsrtnat} %\bibliographystyle{apalike} \end{document}