FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes

Benchmark Model Rank Results
image-classification-on-cifar-10FlexTCN-7#175Percentage correct: 92.2
sequential-image-classification-on-noiseFlexTCN-6#1% Test Accuracy: 69.87%
sequential-image-classification-on-sequentialFlexTCN-4#3Permuted Accuracy: 98.72%
sequential-image-classification-on-sequentialFlexTCN-6#28Unpermuted Accuracy: 99.62%
sequential-image-classification-on-sequential-1FlexTCN-6#6Unpermuted Accuracy: 80.82%
time-series-on-speech-commandsFlexTCN-4#3% Test Accuracy: 97.73
time-series-on-speech-commandsFlexTCN-6#6% Test Accuracy (Raw Data): 91.73