mlsanityData to 7 Oct 2026

UniLION: Towards Unified Autonomous Driving Model with Linear Group RNNs

arXiv 2511.01768 ↗First posted 3 Nov 2025cs.CVcited by 3

On the map: End-to-end autonomous driving → End-to-end system architecture

Abstract (arXiv)

Although transformers have demonstrated remarkable capabilities across various domains, their quadratic attention mechanisms introduce significant computational overhead when processing long-sequence data. In this paper, we present a unified autonomous driving model, UniLION, which efficiently handles large-scale LiDAR point clouds, high-resolution multi-view images, and even temporal sequences based on the linear group RNN operator (i.e., performs linear RNN for grouped features). Remarkably, UniLION serves as a single versatile architecture that can seamlessly support multiple specialized variants (i.e., LiDAR-only, temporal LiDAR, multi-modal, and multi-modal temporal fusion configurations) without requiring explicit temporal or multi-modal fusion modules. Moreover, UniLION consistently delivers competitive and even state-of-the-art performance across a wide range of core tasks, including 3D perception (e.g., 3D object detection, 3D object tracking, 3D occupancy prediction, BEV map segmentation), prediction (e.g., motion prediction), and planning (e.g., end-to-end planning). This unified paradigm naturally simplifies the design of multi-modal and multi-task autonomous driving systems while maintaining superior performance. Ultimately, we hope UniLION offers a fresh perspective on the development of 3D foundation models in autonomous driving. Code is available at https://github.com/happinesslz/UniLION

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Placed in End-to-end system architecture, one of 9 directions in End-to-end autonomous driving. Builds on 92 of its 100 references that are in the map. First posted on arXiv 3 Nov 2025.

Builds on

92 of 100 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024DeepInteraction++: Multi-Modality Interaction for Autonomous Driving9 Aug
  2. 2024SparseDrive: End-to-End Autonomous Driving via Sparse Scene Representation30 May
  3. 2023Is Ego Status All You Need for Open-Loop End-to-End Autonomous Driving?5 Dec
  4. 2023FusionAD: Multi-modality Fusion for Prediction and Planning Tasks of Autonomous Driving2 Aug

Other references · 88

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  4. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,616
  5. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  6. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  7. 2022BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformerscited by 1,213
  8. 2020Center-based 3D Object Detection and Trackingcited by 1,143
  9. 2022Planning-oriented Autonomous Drivingcited by 1,068
  10. 2024Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Modelcited by 1,051
  11. 2020Lift, Splat, Shoot: Encoding Images From Arbitrary Camera Rigs by Implicitly Unprojecting to 3Dcited by 1,040
  12. 2024VMamba: Visual State Space Modelcited by 955
38 more
  1. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 939
  2. 2019PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detectioncited by 890
  3. 2022BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representationcited by 844
  4. 2023RWKV: Reinventing RNNs for the Transformer Eracited by 814
  5. 2021BEVDet: High-performance Multi-camera 3D Object Detection in Bird-Eye-Viewcited by 711
  6. 2019Deep Hough Voting for 3D Object Detection in Point Cloudscited by 648
  7. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  8. 2023VAD: Vectorized Scene Representation for Efficient Autonomous Drivingcited by 613
  9. 2023Retentive Network: A Successor to Transformer for Large Language Modelscited by 594
  10. 2023Gated Linear Attention Transformers with Hardware-Efficient Trainingcited by 490
  11. 2022TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformerscited by 464
  12. 20203DSSD: Point-based 3D Single Stage Object Detectorcited by 440
  13. 2020Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detectioncited by 409
  14. 2023Resurrecting Recurrent Neural Networks for Long Sequencescited by 397
  15. 2019From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Networkcited by 366
  16. 2019Class-balanced Grouping and Sampling for Point Cloud 3D Object Detectioncited by 350
  17. 2022BEVFusion: A Simple and Robust LiDAR-Camera Fusion Frameworkcited by 334
  18. 2024Learning to (Learn at Test Time): RNNs with Expressive Hidden Statescited by 329
  19. 2023Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous Drivingcited by 325
  20. 2024xLSTM: Extended Long Short-Term Memorycited by 320
  21. 2023Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D Object Detectioncited by 313
  22. 2019STD: Sparse-to-Dense 3D Object Detector for Point Cloudcited by 282
  23. 2022Cross-view Transformers for real-time Map-view Semantic Segmentationcited by 264
  24. 2021PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detectioncited by 232
  25. 2024Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language Modelscited by 232
  26. 2023VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Trackingcited by 221
  27. 2018Orthographic Feature Transform for Monocular 3D Object Detectioncited by 201
  28. 2021Embracing Single Stride 3D Object Detector with Sparse Transformercited by 193
  29. 2021Multimodal Virtual Point 3D Detectioncited by 166
  30. 2024PointMamba: A Simple State Space Model for Point Cloud Analysiscited by 166
  31. 2020EPNet: Enhancing Point Features with Image Semantics for 3D Object Detectioncited by 164
  32. 2023FB-OCC: 3D Occupancy Prediction based on Forward-Backward View Transformationcited by 162
  33. 2022M2BEV: Multi-Camera Joint 3D Detection and Segmentation with Unified Birds-Eye View Representationcited by 155
  34. 2024Eagle and Finch: RWKV with Matrix-Valued States and Dynamic Recurrencecited by 155
  35. 2020PnPNet: End-to-End Perception and Prediction with Tracking in the Loopcited by 143
  36. 2022Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Cloudscited by 131
  37. 2023RenderOcc: Vision-Centric 3D Occupancy Prediction with 2D Rendering Supervisioncited by 126
  38. 2022DeepInteraction: 3D Object Detection via Modality Interactioncited by 125

50 most cited of 88 All 88 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 1.

3 papers.

  1. 2025DrivePI: Spatial-aware 4D MLLM for Unified Autonomous Driving Understanding, Perception, Prediction and Planningcited by 15
  2. 2025GeoTeacher: Geometry-Guided Semi-Supervised 3D Object Detectioncited by 1
  3. 2026FlowR2A: Learning Reward-to-Action Distribution for Multimodal Driving Planningcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2026UniDriveDreamer: A Single-Stage Multimodal World Model for Autonomous Driving
  2. 2023DeepIPCv2: LiDAR-powered Robust Environmental Perception and Navigational Control for Autonomous Vehicle
  3. 2026123D: Unifying Multi-Modal Autonomous Driving Data at Scale
  4. 2021LiDAR Cluster First and Camera Inference Later: A New Perspective Towards Autonomous Driving
  5. 2025PRIX: Learning to Plan from Raw Pixels for End-to-End Autonomous Driving
  6. 2025DriveLiDAR4D: Sequential and Controllable LiDAR Scene Generation for Autonomous Driving
  7. 2021Efficient and Robust LiDAR-Based End-to-End Navigation
  8. 2025DepthVision: Enabling Robust Vision-Language Models with GAN-Based LiDAR-to-RGB Synthesis for Autonomous Driving