FLASHSWIN: Unlocking Large Windows and Dense Tokens in Swin Vision Transformers with Memory Efficient Attention

arXiv:2610.04664 · cs.CV · Submitted 2026-10-03 · Read on arXiv

cs.CV

Submitted: 2026-10-03

Updated: 2026-10-03

Code: https://github.com/tushaarkataria/FlashSwin

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Transformer — 1× in this paper · explained in 2 episodes

Sources

• Token Merging: Your ViT But Faster
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• Training Deep Nets with Sublinear Memory Cost
• Vision Transformer Adapter for Dense Predictions
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Benchmarking Neural Network Robustness to Common Corruptions and Perturbations
• ImageNet-21K Pretraining for the Masses
• DINOv3
• Focal Self-attention for Local-Global Interactions in Vision Transformers

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models