Papers › P2T: Pyramid Pooling Transformer for Scene Understanding

P2T: Pyramid Pooling Transformer for Scene Understanding

22 Jun 2021arXiv:2106.12011archive 2025-07-28

Yu-Huan Wu, Yun Liu, Xin Zhan, Ming-Ming Cheng

Recently, the vision transformer has achieved great success by pushing the state-of-the-art of various vision tasks. One of the most challenging problems in the vision transformer is that the large sequence length of image tokens leads to high computational cost (quadratic complexity). A popular solution to this problem is to use a single pooling operation to reduce the sequence length. This paper considers how to improve existing vision transformers, where the pooled feature extracted by a single pooling operation seems less powerful. To this end, we note that pyramid pooling has been demonstrated to be effective in various vision tasks owing to its powerful ability in context abstraction. However, pyramid pooling has not been explored in backbone network design. To bridge this gap, we propose to adapt pyramid pooling to Multi-Head Self-Attention (MHSA) in the vision transformer, simultaneously reducing the sequence length and capturing powerful contextual features. Plugged with our pooling-based MHSA, we build a universal vision transformer backbone, dubbed Pyramid Pooling Transformer (P2T). Extensive experiments demonstrate that, when applied P2T as the backbone network, it shows substantial superiority in various vision tasks such as image classification, semantic segmentation, object detection, and instance segmentation, compared to previous CNN- and transformer-based networks. The code will be released at https://github.com/yuhuan-wu/P2T.

Paper PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

yuhuan-wu/P2T officialmentioned in paperpytorch report

yuhuan-wu/EDN mentioned on GitHubpytorch report

yuhuan-wu/RDPNet mentioned on GitHubpytorch report

yuhuan-wu/mobilesal mentioned 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 Instance Segmentation Object Detection RGB Salient Object Detection Saliency Detection Scene Understanding Semantic Segmentation image-classification object-detection

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
RGB Salient Object Detection	DUTS-TE	P2T-Small	MAE	0.029	#17 of 31	Archive leaderboard	report
RGB Salient Object Detection	DUTS-TE	P2T-Small	max F-measure	0.912	#17 of 31	Archive leaderboard	report
RGB Salient Object Detection	DUTS-TE	P2T-Tiny	MAE	0.033	#18 of 31	Archive leaderboard	report
RGB Salient Object Detection	DUTS-TE	P2T-Tiny	max F-measure	0.895	#18 of 31	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 Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer Vision 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