List your company

5-way Few-shot Classification on Omniglot (test)

99.8Accuracy (1-shot)DCN4
82.1286.7191.395.89Mar 15, 2017Aug 16, 2017Jan 17, 2018Jun 20, 2018Nov 21, 2018Apr 24, 2019Sep 25, 2019

Evaluation results

MethodLinks
DCN4layers=42019.0999.899.891——
Relation nets2019.0999.699.8——
Meta-SGD2019.0999.5399.93——
MAML++Ways=52018.1099.4799.93——
Support nets6layers=62019.0999.3799.8——
Support nets4layers=42019.0999.2499.75——
SNAIL2019.0999.0799.78——
MetaNet2017.1298.95———
Meta-NetworksWays=52018.1098.95———
Prototypical NetworksDistance=Euclidean, Fine-tune=No2017.0398.899.7——
Prototypical nets2019.0998.899.7——
MAML2019.0998.799.9——
MAML (local replication)Ways=52018.1098.5799.82——
Memory Mod.Ways=52018.1098.499.6——
Matching NetworksDistance=Cosine, Fine-tune=No2017.0398.198.9——
Neural StatisticianFine-tune=No2017.0398.199.5——
Matching Nets2017.1298.198.9——
Matching NetsWays=52018.1098.198.9——
Neural StatisticianWays=52018.1098.199.5——
Matching nets2019.0998.198.9——
Neural statistician2019.0998.199.5——
Matching NetworksDistance=Cosine, Fine-tune=Yes2017.0397.998.7——
Siamese Net2017.1297.398.4——
Siamese NetsWays=52018.1097.398.4——
Siamese nets2019.0997.398.4——
MANN2017.1282.894.9——
MANN2019.0982.894.9——
APLShots=12019.08——97.9—
APLShots=52019.08——99.9—
DLMShots=12019.08——98.8—
DLMShots=52019.08——95.4—
Few-Shot Learning with Global Class RepresentationsShots=12019.08——99.720.06
Few-Shot Learning with Global Class RepresentationsShots=52019.08——99.90.1
MAShots=12019.08——98.70.4
MAShots=52019.08——99.90.1
MGShots=12019.08——99.670.18
MGShots=52019.08——99.860.11
MMNShots=12019.08——99.280.08
MMNShots=52019.08——99.770.04
MNShots=12019.08——97.9—
MNShots=52019.08——98.7—
PNShots=12019.08——98.8—
PNShots=52019.08——99.7—
RNShots=12019.08——99.60.2
RNShots=52019.08——99.80.1