baseline:LogoSense

Step1:

• Download the code for all baselines: https://drive.google.com/drive/folders/1YpKR_Nye4E11FCbPbePAAJG4UcqkIsfZ?usp=sharing

• Navigate to LogoSense

• Download Sampled_phish_5brand, and benign_sample_30k dataset, put them into same directory

Step2: Run

python test.py -f Sampled_phish_5brand -m phish -md SVM_imageset -ts threshold

Or

python test.py -f benign_sample_30k -m benign -md SVM_imageset -ts threshold

Step3:

• The format of result will be:

Sampled_phish_5brand/DHL Airways, Inc.+2019-07-29-00`14`45/shot.png DHL Airways, Inc [1.12441876] 23.74169397354126

Sampled_phish_5brand/PayPal Inc.+2020-07-04-17`32`34/shot.png 0 24.189546585083008

Sampled_phish_5brand/PayPal Inc.+2020-05-31-11`39`07/shot.png 0 23.52459716796875

• First column is path, second column is predicted target, third column is the decision function value, last column is total runtime

** The code is forked from https://github.com/SamPlvs/Object-detection-via-HOG-SVM