iFormer: Inception Transformer
This is a PyTorch implementation of iFormer proposed by our paper "Inception Transformer".
The code and model will be released soon.
| Model | #params | FLOPs | Image resolution | acc@1 | Model |
|---|---|---|---|---|---|
| iFormer-S | 20M | 4.8G | 224 | 83.4 | soon |
| iFormer-B | 48M | 9.4G | 224 | 84.6 | soon |
| iFormer-L | 87M | 14.0G | 224 | 84.8 | soon |
Fine-tuning Results with larger resolution (384x384) on ImageNet-1K
| Model | #params | FLOPs | Image resolution | acc@1 | Model |
|---|---|---|---|---|---|
| iFormer-S | 20M | 16.1G | 384 | 84.6 | soon |
| iFormer-B | 48M | 30.5G | 384 | 85.7 | soon |
| iFormer-L | 87M | 45.3G | 384 | 85.8 | soon |
All models are based on Mask R-CNN and trained by 1x training schedule.
| Backbone | #Param. | FLOPs | box mAP | mask mAP |
|---|---|---|---|---|
| iFormer-S | 40M | 263G | 46.2 | 41.9 |
| iFormer-B | 67M | 351G | 48.3 | 43.3 |
| Backbone | Method | #Param. | FLOPs | mIoU |
|---|---|---|---|---|
| iFormer-S | FPN | 24M | 181G | 48.6 |
| iFormer-S | Upernet | 49M | 938G | 48.4 |
Our implementation is mainly based on the following codebases. We gratefully thank the authors for their wonderful works.
pytorch-image-models, mmdetection, mmsegmentation.
Besides, Weihao Yu would like to thank TPU Research Cloud (TRC) program for the support of partial computational resources.