Information
Contact Name: Heechul Jung
Email: heechul@dgist.ac.kr
Organization: DGIST
Method's Name: (C) Joint fine-tuning with DropCNN
Reference: H.Jung, MK Choi, J.Jung, JH Lee, S.Kwon, WY Jung "ResNet-based Vehicle Classification and Localization in Traffic Surveillance Systems", Traffic Surveillance Workshop and Challenge, CVPR 2017.
Project Website:
Evaluation Metrics

Precision of each category :

$$ Pre_i = \frac{TP_i}{TP_i + FP_i}$$


Articulated Truck Bicycle Bus Car Motorcycle Non-motorized Vehicle Pedestrian Pickup Truck Single Unit Truck Work Van Background
0.9213 0.9376 0.9902 0.9871 0.9869 0.9582 0.9710 0.9231 0.8395 0.9733 0.9945

Recall of each category:

$$ Rec_i = \frac{TP_i}{TP_i + FN_i}$$


Articulated Truck Bicycle Bus Car Motorcycle Non-motorized Vehicle Pedestrian Pickup Truck Single Unit Truck Work Van Background
0.9324 0.8949 0.9779 0.9853 0.9111 0.5228 0.9406 0.9539 0.8336 0.9166 0.9984

Accuracy: 0.9795

$$ Acc = \frac{TP}{Number of Testing Images}$$


Mean Recall: 0.8970

$$ {mRec} = {\rm mean}({Rec_i})$$


Mean Precision: 0.9530

$$ {mPre} = {\rm mean}({Pre_i})$$


Cohen Kappa Score: 0.9681

Confusion Matrix (%):

Predicted
Articulated Truck Bicycle Bus Car Motorcycle Non-motorized Vehicle Pedestrian Pickup Truck Single Unit Truck Work Van Background
True Articulated Truck 93.24 0.00 0.35 0.93 0.00 0.23 0.00 0.23 4.33 0.15 0.54
Bicycle 0.18 89.49 0.18 2.45 0.70 0.00 6.48 0.00 0.00 0.00 0.53
Bus 0.23 0.00 97.79 1.47 0.00 0.00 0.00 0.35 0.08 0.08 0.00
Car 0.00 0.00 0.00 98.53 0.00 0.00 0.00 1.42 0.00 0.04 0.00
Motorcycle 0.00 1.82 0.00 2.22 91.11 0.00 0.40 0.00 0.00 0.00 4.44
Non-motorized Vehicle 10.73 0.23 0.23 0.91 0.00 52.28 0.68 3.65 13.01 2.05 16.21
Pedestrian 0.00 1.47 0.00 0.26 0.06 0.00 94.06 0.00 0.00 0.00 4.15
Pickup Truck 0.00 0.00 0.02 4.50 0.00 0.00 0.00 95.39 0.09 0.01 0.00
Single Unit Truck 10.31 0.00 0.16 0.47 0.00 0.16 0.00 2.58 83.36 1.09 1.88
Work Van 0.08 0.04 0.08 5.62 0.00 0.04 0.04 0.78 0.78 91.66 0.87
Background 0.04 0.00 0.02 0.07 0.00 0.00 0.00 0.01 0.01 0.02 99.84