architectureIntroduced by MobileNetV2 · 2018
Inverted residuals & linear bottlenecks
Expand, filter with depthwise conv, then project back down to thin residual bottlenecks.
Drafted by AI · not yet reviewed
How this idea evolved
Suggested from citations. No curator has recorded what this idea builds on yet. These ideas from the same theme come from papers that MobileNetV2 cites, directly or one step removed. Citation is a fact; the connection between the ideas is not verified.
Factor a convolution into per-channel and 1×1 steps to cut computation drastically.
Search for a network's building block automatically instead of designing it by hand.
Papers using this
- 2018MnasNet