# FlatFormer: Flattened Window Attention for Efficient Point Cloud Transformer

arXiv: https://arxiv.org/abs/2301.08739 · HTML: https://mlsanity.com/paper/2301.08739/

## Key facts

- Cited by 73 papers in the map (data through 14 Oct 2026).
- Builds on 93 of its 102 references that are in the map.
- First posted on arXiv 20 Jan 2023.

## Builds on (most cited first, 30 of 93)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [End-to-End Object Detection with Transformers](https://mlsanity.com/paper/2005.12872/) (2020)
- [PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation](https://mlsanity.com/paper/1612.00593/) (2016)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space](https://mlsanity.com/paper/1706.02413/) (2017)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://mlsanity.com/paper/2010.04159/) (2020)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [Scalability in Perception for Autonomous Driving: Waymo Open Dataset](https://mlsanity.com/paper/1912.04838/) (2019)
- [Dynamic Graph CNN for Learning on Point Clouds](https://mlsanity.com/paper/1801.07829/) (2018)
- [Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention](https://mlsanity.com/paper/2006.16236/) (2020)
- [Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions](https://mlsanity.com/paper/2102.12122/) (2021)
- [PointPillars: Fast Encoders for Object Detection from Point Clouds](https://mlsanity.com/paper/1812.05784/) (2018)
- [3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation](https://mlsanity.com/paper/1606.06650/) (2016)
- [Point Transformer](https://mlsanity.com/paper/2011.00931/) (2020)
- [Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers](https://mlsanity.com/paper/2012.15840/) (2020)
- [VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection](https://mlsanity.com/paper/1711.06396/) (2017)
- [Swin Transformer V2: Scaling Up Capacity and Resolution](https://mlsanity.com/paper/2111.09883/) (2021)
- [BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers](https://mlsanity.com/paper/2203.17270/) (2022)
- [4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks](https://mlsanity.com/paper/1904.08755/) (2019)
- [KPConv: Flexible and Deformable Convolution for Point Clouds](https://mlsanity.com/paper/1904.08889/) (2019)
- [Per-Pixel Classification is Not All You Need for Semantic Segmentation](https://mlsanity.com/paper/2107.06278/) (2021)
- [Center-based 3D Object Detection and Tracking](https://mlsanity.com/paper/2006.11275/) (2020)
- [Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet](https://mlsanity.com/paper/2101.11986/) (2021)
- [Transformers in Vision: A Survey](https://mlsanity.com/paper/2101.01169/) (2021)

## Built on by (most cited first, 30 of 73)

- [Point Transformer V3: Simpler, Faster, Stronger](https://mlsanity.com/paper/2312.10035/) (2023)
- [UniTR: A Unified and Efficient Multi-Modal Transformer for Bird's-Eye-View Representation](https://mlsanity.com/paper/2308.07732/) (2023)
- [Point Cloud Mamba: Point Cloud Learning via State Space Model](https://mlsanity.com/paper/2403.00762/) (2024)
- [Voxel Mamba: Group-Free State Space Models for Point Cloud based 3D Object Detection](https://mlsanity.com/paper/2406.10700/) (2024)
- [LION: Linear Group RNN for 3D Object Detection in Point Clouds](https://mlsanity.com/paper/2407.18232/) (2024)
- [TorchSparse++: Efficient Training and Inference Framework for Sparse Convolution on GPUs](https://mlsanity.com/paper/2311.12862/) (2023)
- [FSD V2: Improving Fully Sparse 3D Object Detection with Virtual Voxels](https://mlsanity.com/paper/2308.03755/) (2023)
- [Fully Sparse Fusion for 3D Object Detection](https://mlsanity.com/paper/2304.12310/) (2023)
- [Once Detected, Never Lost: Surpassing Human Performance in Offline LiDAR based 3D Object Detection](https://mlsanity.com/paper/2304.12315/) (2023)
- [RadarPillars: Efficient Object Detection from 4D Radar Point Clouds](https://mlsanity.com/paper/2408.05020/) (2024)
- [Erwin: A Tree-based Hierarchical Transformer for Large-scale Physical Systems](https://mlsanity.com/paper/2502.17019/) (2025)
- [SEED: A Simple and Effective 3D DETR in Point Clouds](https://mlsanity.com/paper/2407.10749/) (2024)
- [LitePT: Lighter Yet Stronger Point Transformer](https://mlsanity.com/paper/2512.13689/) (2025)
- [ConDaFormer: Disassembled Transformer with Local Structure Enhancement for 3D Point Cloud Understanding](https://mlsanity.com/paper/2312.11112/) (2023)
- [Locality-Sensitive Hashing-Based Efficient Point Transformer with Applications in High-Energy Physics](https://mlsanity.com/paper/2402.12535/) (2024)
- [LiDARFormer: A Unified Transformer-based Multi-task Network for LiDAR Perception](https://mlsanity.com/paper/2303.12194/) (2023)
- [ScatterFormer: Efficient Voxel Transformer with Scattered Linear Attention](https://mlsanity.com/paper/2401.00912/) (2024)
- [V2X-PC: Vehicle-to-everything Collaborative Perception via Point Cluster](https://mlsanity.com/paper/2403.16635/) (2024)
- [On permutation-invariant neural networks](https://mlsanity.com/paper/2403.17410/) (2024)
- [UniMamba: Unified Spatial-Channel Representation Learning with Group-Efficient Mamba for LiDAR-based 3D Object Detection](https://mlsanity.com/paper/2503.12009/) (2025)
- [MixSup: Mixed-grained Supervision for Label-efficient LiDAR-based 3D Object Detection](https://mlsanity.com/paper/2401.16305/) (2024)
- [A Point-Based Approach to Efficient LiDAR Multi-Task Perception](https://mlsanity.com/paper/2404.12798/) (2024)
- [FlatFusion: Delving into Details of Sparse Transformer-based Camera-LiDAR Fusion for Autonomous Driving](https://mlsanity.com/paper/2408.06832/) (2024)
- [FSHNet: Fully Sparse Hybrid Network for 3D Object Detection](https://mlsanity.com/paper/2506.03714/) (2025)
- [SparseFusion: Efficient Sparse Multi-Modal Fusion Framework for Long-Range 3D Perception](https://mlsanity.com/paper/2403.10036/) (2024)
- [HOTFormerLoc: Hierarchical Octree Transformer for Versatile Lidar Place Recognition Across Ground and Aerial Views](https://mlsanity.com/paper/2503.08140/) (2025)
- [Towards Stable 3D Object Detection](https://mlsanity.com/paper/2407.04305/) (2024)
- [BFANet: Revisiting 3D Semantic Segmentation with Boundary Feature Analysis](https://mlsanity.com/paper/2503.12539/) (2025)
- [How Many Tokens Do 3D Point Cloud Transformer Architectures Really Need?](https://mlsanity.com/paper/2511.05449/) (2025)
- [Sparse Refinement for Efficient High-Resolution Semantic Segmentation](https://mlsanity.com/paper/2407.19014/) (2024)

## Abstract (arXiv)

Transformer, as an alternative to CNN, has been proven effective in many modalities (e.g., texts and images). For 3D point cloud transformers, existing efforts focus primarily on pushing their accuracy to the state-of-the-art level. However, their latency lags behind sparse convolution-based models (3x slower), hindering their usage in resource-constrained, latency-sensitive applications (such as autonomous driving). This inefficiency comes from point clouds' sparse and irregular nature, whereas transformers are designed for dense, regular workloads. This paper presents FlatFormer to close this latency gap by trading spatial proximity for better computational regularity. We first flatten the point cloud with window-based sorting and partition points into groups of equal sizes rather than windows of equal shapes. This effectively avoids expensive structuring and padding overheads. We then apply self-attention within groups to extract local features, alternate sorting axis to gather features from different directions, and shift windows to exchange features across groups. FlatFormer delivers state-of-the-art accuracy on Waymo Open Dataset with 4.6x speedup over (transformer-based) SST and 1.4x speedup over (sparse convolutional) CenterPoint. This is the first point cloud transformer that achieves real-time performance on edge GPUs and is faster than sparse convolutional methods while achieving on-par or even superior accuracy on large-scale benchmarks.