aboutsummaryrefslogtreecommitdiff
usage: evaluate.py [-h] [-t] [-c] [-k] [-m] [-e] [-r] [-a RERANKA]
                   [-b RERANKB] [-l RERANKL] [-n NEIGHBORS] [-v] [-s SHOWRANK]
                   [-1] [-2] [-M MULTRANK] [-C] [--data DATA] [-K KMEAN] [-A]
                   [-P PCA]

optional arguments:
  -h, --help            show this help message and exit
  -t, --train           Use train data instead of query and gallery
  -c, --conf_mat        Show visual confusion matrix
  -k, --kmean_alt       Perform clustering with generalized labels(not actual
                        kmean)
  -m, --mahalanobis     Perform Mahalanobis Distance metric
  -e, --euclidean       Use standard euclidean distance
  -r, --rerank          Use k-reciprocal rernaking
  -a RERANKA, --reranka RERANKA
                        Parameter k1 for rerank
  -b RERANKB, --rerankb RERANKB
                        Parameter k2 for rerank
  -l RERANKL, --rerankl RERANKL
                        Parameter lambda for rerank
  -n NEIGHBORS, --neighbors NEIGHBORS
                        Use customized ranklist size NEIGHBORS
  -v, --verbose         Use verbose output
  -s SHOWRANK, --showrank SHOWRANK
                        Save ranklist pics id in a txt file for first SHOWRANK
                        queries
  -1, --normalise       Normalise features
  -2, --standardise     Standardise features
  -M MULTRANK, --multrank MULTRANK
                        Run for different ranklist sizes equal to MULTRANK
  -C, --comparison      Compare baseline and improved metric
  --data DATA           Folder containing data
  -K KMEAN, --kmean KMEAN
                        Perform Kmean clustering, KMEAN number of clusters
  -A, --mAP             Display Mean Average Precision
  -P PCA, --PCA PCA     Perform pca with PCA eigenvectors

EXAMPLES 1 for evaluate.py:

EXAMPLE 1.1: Run euclidean distance with top n

evaluate.py -e -n 10

or simply

evaluate.py -n 10

EXAMPLE 1.2: Run euclidean distance for the first 10 values of top n and graph them

evaluate.py -M 10

EXAMPLE 1.3: Run comparison between baseline and rerank for the first 5 values of top n and graph them

evaluate.py -M 5 -C

EXAMPLE 1.4: Run for kmeans, 10 clusters

evaluate.py -K 10

EXAMPLE 1.5: Run for mahalanobis, using PCA for top 100 eigenvectors to speed up the calculation

evaluate.py -m -P 100

EXAMPLE 1.6: Run rerank for customized values of RERANKA, RERANKB and RERANKL

evaluate.py -r -a 11 -b 3 -l 0.3

EXAMPLE 1.7: Run on the training set with euclidean distance and normalize feature vectors. Draw confusion matrix at the end

evaluate.py -t -1 -c

EXAMPLE 1.8: Run euclidean distance standardising the feature data for the first 10 values of top n and graph them

evaluate.py -2 -M 10

EXAMPLE 1.9: Run for rerank top 10 and save the names of the images that compose the ranklist for the first 5 queries: query.txt, ranklist.txt

evaluate.py -r -s 5 -n 10

EXAMPLE 1.10: Display mAP. It is advisable to use high n to obtain an accurate result

evaluate.py -A -n 5000

EXAMPLE 1.11: Run euclidean distance specifying a different data folder location

for data int the same folder as evaluate.py:

evaluate.py --data ./

or for data in another folder:

evaluate.py --data ./foo/bar/

EXAMPLES 2 for opt.py:

EXAMPLE 2.1: optimize top 1 accuracy for k1, k2, lambda speeding up the process with PCA, top 50 eigenvectors

opt.py -P 50

EXAMPLE 2.2: optimize mAP for k1, k2, lambda speeding up the process with PCA, top 50 eigenvectors

opt.py -P 50 -A