Sequencer: Deep LSTM for Image Classification

Open paper
Benchmark Model Rank Results
domain-generalization-on-imagenet-aSequencer2D-L#24Top-1 accuracy %: 35.5
domain-generalization-on-imagenet-cSequencer2D-L#25mean Corruption Error (mCE): 48.9
domain-generalization-on-imagenet-rSequencer2D-L#26Top-1 Error Rate: 51.9
domain-generalization-on-imagenet-sketchSequencer2D-L#19Top-1 accuracy: 35.8
image-classification-on-imagenetSequencer2D-L↑392#297Top 1 Accuracy: 84.6%Number of params: 54MGFLOPs: 50.7
image-classification-on-imagenetSequencer2D-L#418Top 1 Accuracy: 83.4%Number of params: 54MGFLOPs: 16.6
image-classification-on-imagenetSequencer2D-M#485Top 1 Accuracy: 82.8%Number of params: 38MGFLOPs: 11.1
image-classification-on-imagenetSequencer2D-S#537Top 1 Accuracy: 82.3%Number of params: 28MGFLOPs: 8.4
image-classification-on-imagenet-realSequencer2D-L#29Accuracy: 87.9
image-classification-on-imagenet-v2Sequencer2D-L#21Top 1 Accuracy: 73.4

Rank counts only results with a code link.