Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Spherical MNIST

Benchmarks

Task NameDataset NameSOTA ResultTrend
ClassificationSpherical MNIST rotated level-4 mesh (train and test (R/R))
Accuracy99.37
16
Image ClassificationSpherical MNIST
Accuracy99.36
13
Image ClassificationSpherical MNIST NR/NR
Accuracy99.45
12
Digit ClassificationSpherical MNIST (test)
Accuracy98.6
9
MNIST ClassificationSpherical MNIST Randomly Rotated
Accuracy98.1
7
MNIST ClassificationSpherical MNIST non-rotated
Accuracy98.7
7
Image ClassificationSpherical MNIST NR/R
Accuracy99.08
5
Image ClassificationSpherical MNIST 64x64 resolution (test)
Accuracy99.31
4
Image ClassificationSpherical MNIST 256x256 resolution (test)
Accuracy99.29
3
Image ClassificationSpherical MNIST 128x128 resolution (test)
Accuracy99.32
3
Image ClassificationSpherical MNIST 32x32 resolution (test)
Accuracy99.28
3
Vector-to-label digit classificationSpherical MNIST Vector (test)
NR/NR Score99.45
3
Scalar-to-label digit classificationSpherical MNIST Scalar (test)
NR/NR Rate99.47
3
ClassificationSpherical MNIST non-rotated train and rotated test (N/R) level-4 mesh
Accuracy69.99
3
ClassificationSpherical MNIST non-rotated level-4 mesh (train test)
Accuracy99.45
3
Digit RecognitionSpherical MNIST (test)
Accuracy99.23
3
Showing 16 of 16 rows