Fourier or Wavelet bases as counterpart self-attention in spikformer for efficient visual classification
Open paper
Benchmark
Model
Rank
Results
image-classification-on-cifar-10
FWformer-4-384 (2D-WT-Haar, MS)
–
Percentage correct: 95.6
image-classification-on-cifar-100
FWformer-4-384 (2D-FFT, MS)
–
Percentage correct: 78.3
Rank counts only results with a code link.