Increasing Trustworthiness of Deep Neural Networks via Accuracy Monitoring

Open paper
Benchmark Model Rank Results
image-classification-on-stl-10Accuracy Monitoring–Percentage correct: 68.62
image-classification-on-stl-10Entropy–Percentage correct: 71.65
image-classification-on-stl-10MP–Percentage correct: 71.05
image-classification-on-stl-10MP*–Percentage correct: 93.19
image-classification-on-stl-10TS–Percentage correct: 88.03

Rank counts only results with a code link.