From 95f25f18f6af9a2faa08ea288c91bb7d21a4eb5b Mon Sep 17 00:00:00 2001 From: nunzip Date: Thu, 13 Dec 2018 20:21:19 +0000 Subject: Add data variance info, correct grammar --- report/paper.md | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/report/paper.md b/report/paper.md index b9aa477..73c49de 100755 --- a/report/paper.md +++ b/report/paper.md @@ -38,7 +38,7 @@ of its nearest neighbor(s). The distance between images can be calculated throug different distance metrics, however one of the most commonly used is euclidean distance: -$$ \textrm{NN}(x) = \operatorname*{argmin}_{i\in[m]} \|x-x_i\| $$ +$$ \textrm{NN}(x) = \operatorname*{argmin}_{i\in[m]} \|x-x_i\|. $$ # Baseline Evaluation @@ -73,7 +73,7 @@ Applying magnitude normalization (scaling feature vectors to unit length) had a effect re-identification. Furthemore standartization by removing feature mean and deviation also had negative effect on performance as seen on figure \ref{fig:baselineacc}. This may be due to the fact that we are removing feature scaling that was introduced by the Neural network, -such that some of the features are more significant than the others. By standartizing our +such that some of the features are more significant than others. By standartizing our features at this point, we remove such scaling and may be losing using useful metrics. ## kMeans Clustering @@ -122,7 +122,8 @@ $$ d_M(p,g_i) = (p-g_i)^TM(p-g_i). $$ When performing Mahalanobis with the covariance matrix $M$ generated from the training set, reported accuracy is reduced to **38%** . -We also attempted to perform the same Mahalanobis metric on a reduced PCA featureset. This allowed for significant execution +We also attempted to perform the same Mahalanobis metric on a reduced PCA featureset. PCA performed with the top 100 eigenvectors +reduces the feature space while keeping 94% of the data variance. This allowed for significant execution time improvements due to the greatly reduced computation requierments for smaller featurespace, but nevertheless demonstrated no improvements over an euclidean metric. @@ -161,7 +162,7 @@ as position, illumination and foreign objects. $R^*(p,k)$ is used to recalculate the distance between query and gallery images. Jaccard metric of the $k$-reciprocal sets is used to calculate the distance -between $p$ and $g_i$ as: $$d_J(p,g_i)=1-\frac{|R^*(p,k)\cap R^*(g_i,k)|}{|R^*(p,k)\cup R^*(g_i,k)|}$$. +between $p$ and $g_i$ as: $$d_J(p,g_i)=1-\frac{|R^*(p,k)\cap R^*(g_i,k)|}{|R^*(p,k)\cup R^*(g_i,k)|}.$$ However, since the neighbors of the query $p$ are close to $g_i$ as well, they would be more likely to be identified as true positive. This implies @@ -255,8 +256,8 @@ The difference between the top $k$ accuracies of the two methods gets smaller as The improved results due to $k$-reciprocal re-ranking may be explained by considering that re-ranks based on second order neighbours, that is, the neighbours of the neighbours. For neighbours which display identifiable features, such as a backpack or binder that is not visible in the query but visible in a close neighbour, the reranking algorithm is able to infer that the strong relationship based on this newly introduced -feature such as a backpack or folder by the neighbour, uniuqly identify other identities in the gallery with the same feature, and moving them up the rankinlist -as a result despite the identifiable feature being hidden in the query. EXAMPLE BACKPACK HERE +feature such as a backpack or folder by the neighbour, uniquely identify other identities in the gallery with the same feature, and moving them up the rankinlist +as a result despite the identifiable feature being hidden in the query. An example of this can be seen in figure \ref{fig:rerank}. \begin{figure} \begin{center} -- cgit v1.2.3-54-g00ecf