Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation

About

Although vision transformers (ViTs) have achieved great success in computer vision, the heavy computational cost hampers their applications to dense prediction tasks such as semantic segmentation on mobile devices. In this paper, we present a mobile-friendly architecture named \textbf{To}ken \textbf{P}yramid Vision Trans\textbf{former} (\textbf{TopFormer}). The proposed \textbf{TopFormer} takes Tokens from various scales as input to produce scale-aware semantic features, which are then injected into the corresponding tokens to augment the representation. Experimental results demonstrate that our method significantly outperforms CNN- and ViT-based networks across several semantic segmentation datasets and achieves a good trade-off between accuracy and latency. On the ADE20K dataset, TopFormer achieves 5\% higher accuracy in mIoU than MobileNetV3 with lower latency on an ARM-based mobile device. Furthermore, the tiny version of TopFormer achieves real-time inference on an ARM-based mobile device with competitive results. The code and models are available at: https://github.com/hustvl/TopFormer

Wenqiang Zhang, Zilong Huang, Guozhong Luo, Tao Chen, Xinggang Wang, Wenyu Liu, Gang Yu, Chunhua Shen• 2022

Related benchmarks

TaskDatasetResultRank
Semantic segmentationADE20K (val)
mIoU39.2
3069
Object DetectionCOCO 2017 (val)
AP31.6
2843
Semantic segmentationPascal Context (test)--
223
Semantic segmentationCoco-Stuff (test)
mIoU33.43
216
Semantic segmentationCamVid
mIoU63.1
82
Anatomy SegmentationRadGenome Anatomy
Dice55.66
19
Semantic segmentationSynthiaSF
mIoU28.37
12
Semantic segmentationCityscapes
mIoU32.76
12
Semantic segmentationCityscapes
mIoU22.15
12
Semantic segmentationCARLA ADV
mIoU32.32
11
Showing 10 of 11 rows

Other info

Code

Follow for update