Papers › IncepFormer: Efficient Inception Transformer with Pyramid Pooling for Semantic Segmentation

IncepFormer: Efficient Inception Transformer with Pyramid Pooling for Semantic Segmentation

6 Dec 2022arXiv:2212.03035archive 2025-07-28

Lihua Fu, Haoyue Tian, Xiangping Bryce Zhai, Pan Gao, Xiaojiang Peng

Semantic segmentation usually benefits from global contexts, fine localisation information, multi-scale features, etc. To advance Transformer-based segmenters with these aspects, we present a simple yet powerful semantic segmentation architecture, termed as IncepFormer. IncepFormer has two critical contributions as following. First, it introduces a novel pyramid structured Transformer encoder which harvests global context and fine localisation features simultaneously. These features are concatenated and fed into a convolution layer for final per-pixel prediction. Second, IncepFormer integrates an Inception-like architecture with depth-wise convolutions, and a light-weight feed-forward module in each self-attention layer, efficiently obtaining rich local multi-scale object features. Extensive experiments on five benchmarks show that our IncepFormer is superior to state-of-the-art methods in both accuracy and speed, e.g., 1) our IncepFormer-S achieves 47.7% mIoU on ADE20K which outperforms the existing best method by 1% while only costs half parameters and fewer FLOPs. 2) Our IncepFormer-B finally achieves 82.0% mIoU on Cityscapes dataset with 39.6M parameters. Code is available:github.com/shendu0321/IncepFormer.

Paper PDF Code

Code

shendu0321/incepformer officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Image Classification Semantic Segmentation

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Image Classification	ImageNet	IPT-B	GFLOPs	7.8	#413 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-B	Number of params	39.3M	#413 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-B	Top 1 Accuracy	83.6%	#413 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-S	GFLOPs	4.7	#486 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-S	Number of params	24.3M	#486 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-S	Top 1 Accuracy	82.9%	#486 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-T	GFLOPs	2.3	#698 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-T	Number of params	14.0M	#698 of 1060	Archive leaderboard	report
Image Classification	ImageNet	IPT-T	Top 1 Accuracy	80.5%	#698 of 1060	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position Encodings Adam Attention BPE Convolution Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections