Scaling the Scattering Transform: Deep Hybrid Networks

Open paper
Benchmark Model Rank Results
image-classification-on-stl-10Scat + WRN 20-8#50Percentage correct: 76.6
image-classification-on-stl-10Exemplar CNN#51Percentage correct: 75.7
image-classification-on-stl-10Stacked what-where AE#52Percentage correct: 74.33
image-classification-on-stl-10CNN#60Percentage correct: 70.7
image-classification-on-stl-10Hierarchical Matching Pursuit (HMP)#63Percentage correct: 64.6
image-classification-on-stl-10Convolutional K-means Network#65Percentage correct: 60.2

Rank counts only results with a code link.