FER2013 Mobilenet Small Benchmark

fer2013_mobilenet_small.txt

Benchmark model: mobilenet_small
Benchmark for non quantized model

LOAD DATASET: START
LOAD DATASET: DONE
INFERENCE: START
INFERENCE: DONE

Angry: Correct cases: 271 / 491, Accuracy: 55.19
Disgust: Correct cases: 0 / 55, Accuracy: 0.00
Fear: Correct cases: 196 / 528, Accuracy: 37.12
Happy: Correct cases: 719 / 879, Accuracy: 81.80
Sad: Correct cases: 240 / 594, Accuracy: 40.40
Surprise: Correct cases: 299 / 416, Accuracy: 71.88
Neutral: Correct cases: 412 / 626, Accuracy: 65.81
All category: Correct cases: 2137 / 3589, Accuracy: 59.54
]/[p]][[ppp
0.53 ms for one data

Benchmark for quantized model
LOAD DATASET: START
LOAD DATASET: DONE
INFERENCE: START
INFERENCE: DONE

Angry: Correct cases: 275 / 491, Accuracy: 56.01
Disgust: Correct cases: 0 / 55, Accuracy: 0.00
Fear: Correct cases: 193 / 528, Accuracy: 36.55
Happy: Correct cases: 724 / 879, Accuracy: 82.37
Sad: Correct cases: 233 / 594, Accuracy: 39.23
Surprise: Correct cases: 295 / 416, Accuracy: 70.91
Neutral: Correct cases: 407 / 626, Accuracy: 65.02
All category: Correct cases: 2127 / 3589, Accuracy: 59.26