Depthwise Convolution is All You Need for Learning Multiple Visual Domains

Open paper
Benchmark Model Rank Results
continual-learning-on-visual-domain-decathlon-10-tasksDepthwise Soft Sharing#2decathlon discipline (Score): 3507
continual-learning-on-visual-domain-decathlon-10-tasksDepthwise Sharing#4decathlon discipline (Score): 3234

Rank counts only results with a code link.