Grouped Pointwise Convolutions Reduce Parameters in Convolutional Neural Networks

Open paper
Benchmark Model Rank Results
image-classification-on-oxford-iiit-petskMobileNet V3 Large 16ch#4Accuracy: 60.4PARAMS: 0.36MFLOPS: 81.0M
image-classification-on-cifar-10kEffNet-B0 32ch#196Percentage correct: 93.75
image-classification-on-cifar-10kMobileNet V3 Large 16ch#211Percentage correct: 92.74
image-classification-on-cifar-10kDenseNet-BC L100 12ch#228Percentage correct: 90.83
image-classification-on-cifar-10kMobileNet 16ch#234Percentage correct: 89.81
image-classification-on-cifar-100kMobileNet V3 Large 16ch#206Percentage correct: 71.36PARAMS: 0.52M

Rank counts only results with a code link.