CapsNet

Reported on 5 benchmarks across 3 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision3 results

Image ClassificationonMultiMNIST
Percentage error· 2017-10-26
5.2
SOTA
Dynamic Routing Between Capsules arXiv:1710.09829
Image ClassificationonsmallNORB
Classification Error· 2017-10-26
3.77
best: 0.9 (Heinsen Routing)
SOTA
Dynamic Routing Between Capsules arXiv:1710.09829
Image ClassificationonMNIST
Percentage error· 2017-10-26
0.25
best: 0.13 (Branching/Merging CNN + Homogeneous Vector Capsules)
Dynamic Routing Between Capsules arXiv:1710.09829

Natural Language Processing1 result

Text ClassificationonJigsaw Toxic Comment Classification Dataset
Validation Accuracy
90.44

Methodology1 result

ClassificationonJigsaw Toxic Comment Classification Dataset
Validation Accuracy
90.44