PyTorch implementation of MobileDet backbones introduced in MobileDets: Searching for Object Detection Architectures for Mobile Accelerators. Ross Wightman's timm library has been used for some helper functions and inspiration for syntax style. The following are the main blocks used in the tensorflow implementation and their corresponding blocks in the timm library:
| Tensorflow | timm |
|---|---|
| fused_conv | EdgeResidual |
| inverterted_bottleneck | InvertedResidual |
| inverted_bottleneck_no_expansion | DepthwiseSeparableConv |
tucker-conv package is used for tucker_conv