From 6aeecafd57b3d070dae23d699f8451a35d1a1ef3 Mon Sep 17 00:00:00 2001 From: Vasil Zlatanov Date: Fri, 15 Mar 2019 22:19:36 +0000 Subject: Improved till page 5 --- report/paper.md | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/report/paper.md b/report/paper.md index fd8512f..0ed78df 100644 --- a/report/paper.md +++ b/report/paper.md @@ -179,7 +179,7 @@ tend to collapse to very small regions. \includegraphics[width=8em]{fig/cdcloss1.png} \includegraphics[width=8em]{fig/cdcloss2.png} \includegraphics[width=8em]{fig/cdcloss3.png} -\caption{cDCGAN G-D loss; Left G/D=1; Middle G/D=3; Right G/D=6} +\caption{cDCGAN G-D loss; Left *G/D=1*; Middle $G/D=3$; Right $G/D=6$} \label{fig:cdcloss} \end{center} \end{figure} @@ -245,13 +245,13 @@ Virtual Batch Normalization is a further optimisation technique proposed by Tim ### Dropout -Despite the difficulties in judging differences between G-D losses and image quality, dropout rate seems to have a noticeable effect on accuracy and Inception Score, with a variation of 3.6% between our best and worst dropout cases. Ultimately, judging from the measurements, it is preferable to use a low dropout rate (0.1 seems to be the one that achieves the best results). +Dropout appears to have a noticeable effect on accuracy and Inception Score, with a variation of 3.6% between our best and worst dropout cases. The measurements indicate that it is preferable to use a low dropout rate (0.1 seems to be the one that achieves the best results). ### G-D Balancing on cDCGAN -Despite achieving lower losses oscillation, using G/D=3 to incentivize generator training did not improve the performance of cDCGAN as it is observed from -the Inception Score and testing accuracy. We obtain in fact 5% less test accuracy, meaning that using this technique in our architecture produces on -average lower quality images when compared to our standard cDCGAN. +Despite achieving lower loss oscillation, using *G/D=3* to incentivize generator training did not improve the performance of cDCGAN as meassured by +the Inception Score and testing accuracy. We obtain 5% less test accuracy, meaning that using this technique in our architecture produces on +lower quality images on average when compared to our standard cDCGAN. # Re-training the handwritten digit classifier @@ -262,10 +262,10 @@ average lower quality images when compared to our standard cDCGAN. In this section we analyze the effect of retraining the classification network using a mix of real and generated data, highlighting the benefits of injecting generated samples in the original training set to boost testing accuracy. -As observed in figure \ref{fig:mix1} we performed two experiments for performance evaluation: +As shown in figure \ref{fig:mix1} we performed two experiments for performance evaluation: -* Keeping the same number of training samples while just changing the ratio of real to generated data (55,000 samples in total). -* Keeping the whole training set from MNIST and adding generated samples from cDCGAN. +* Using the same number of training samples while only changing the ratio of real to generated data (55,000 samples in total). +* Using the whole training set from MNIST and adding generated samples from cDCGAN. \begin{figure} \begin{center} @@ -308,8 +308,7 @@ boosted to 92%, making this technique the most successful attempt of improvement \end{figure} Examples of misclassification are displayed in figure \ref{fig:retrain_fail}. It is visible from a cross comparison between these results and the precision-recall -curve displayed in figure \ref{fig:pr-retrain} that the network we trained performs really well for most of the digits, but the low confidence on digit $8$ lowers -the overall performance. +curve displayed in figure \ref{fig:pr-retrain} that the network performs well for most of the digits, but has is brought down by the relatively low precision for the digit 8, lowering the micro-average'd precision. \begin{figure} \begin{center} @@ -513,7 +512,7 @@ $$ L_{\textrm{total}} = \alpha L_{\textrm{LeNet}} + \beta L_{\textrm{generator}} \begin{figure}[H] \begin{center} \includegraphics[width=18em]{fig/clustcollapse.png} -\caption{cDCGAN G/D=6 PCA Embeddings through LeNet (10000 samples per class)} +\caption{cDCGAN *G/D=6* PCA Embeddings through LeNet (10000 samples per class)} \label{fig:clustcollapse} \end{center} \end{figure} -- cgit v1.2.3