mlsanityData to 14 Oct 2026

XFeat: Accelerated Features for Lightweight Image Matching

arXiv 2404.19174 ↗First posted 30 Apr 2024cs.CVcited by 113

Not on the map. Why?

Abstract (arXiv)

We introduce a lightweight and accurate architecture for resource-efficient visual correspondence. Our method, dubbed XFeat (Accelerated Features), revisits fundamental design choices in convolutional neural networks for detecting, extracting, and matching local features. Our new model satisfies a critical need for fast and robust algorithms suitable to resource-limited devices. In particular, accurate image matching requires sufficiently large image resolutions - for this reason, we keep the resolution as large as possible while limiting the number of channels in the network. Besides, our model is designed to offer the choice of matching at the sparse or semi-dense levels, each of which may be more suitable for different downstream applications, such as visual navigation and augmented reality. Our model is the first to offer semi-dense matching efficiently, leveraging a novel match refinement module that relies on coarse local descriptors. XFeat is versatile and hardware-independent, surpassing current deep learning-based local features in speed (up to 5x faster) with comparable or better accuracy, proven in pose estimation and visual localization. We showcase it running in real-time on an inexpensive laptop CPU without specialized hardware optimizations. Code and weights are available at www.verlab.dcc.ufmg.br/descriptors/xfeat_cvpr24.

Key facts. Cited by 113 papers in the map (data through 14 Oct 2026). Builds on 35 of its 47 references that are in the map. First posted on arXiv 30 Apr 2024.

Builds on

35 of 47 references.

References · 35

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  5. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  6. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  7. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  8. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  9. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
  10. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,320
  11. 2017ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devicescited by 1,200
  12. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
23 more
  1. 2016ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Camerascited by 806
  2. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 737
  3. 2023LightGlue: Local Feature Matching at Light Speedcited by 590
  4. 2018From Coarse to Fine: Robust Hierarchical Localization at Large Scalecited by 559
  5. 2020DISK: Learning local features with policy gradientcited by 312
  6. 2017HPatches: A benchmark and evaluation of handcrafted and learned local descriptorscited by 264
  7. 2016Large-Scale Image Retrieval with Attentive Deep Local Featurescited by 238
  8. 2017Working hard to know your neighbor's margins: Local descriptor learning losscited by 212
  9. 2019R2D2: Repeatable and Reliable Detector and Descriptorcited by 207
  10. 2022DKM: Dense Kernelized Feature Matching for Geometry Estimationcited by 177
  11. 2022ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformercited by 171
  12. 2019MAGSAC++, a fast, reliable and accurate robust estimatorcited by 161
  13. 2020ASLFeat: Learning Local Features of Accurate Shape and Localizationcited by 130
  14. 2020Patch2Pix: Epipolar-Guided Pixel-Level Correspondencescited by 129
  15. 2019Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filterscited by 118
  16. 2021PDC-Net+: Enhanced Probabilistic Dense Correspondence Networkcited by 80
  17. 2021ALIKE: Accurate and Lightweight Keypoint Detection and Descriptor Extractioncited by 77
  18. 2023SiLK -- Simple Learned Keypointscited by 54
  19. 2022ZippyPoint: Fast Interest Point Detection, Description, and Matching through Mixed Precision Discretizationcited by 10
  20. 2023LFM-3D: Learnable Feature Matching Across Wide Baselines Using 3D Signalscited by 8
  21. 2021Extracting Deformation-Aware Local Features by Learning to Deformcited by 7
  22. 2022Learning Geodesic-Aware Local Features from RGB-D Imagescited by 5
  23. 2023Yes, we CANN: Constrained Approximate Nearest Neighbors for local feature-based visual localizationcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 12; 2025: 50; 2026: 51.

113 papers.

Sort:

By area:

  1. 2024CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agentscited by 39
  2. 2025On-the-fly Reconstruction for Large-Scale Novel View Synthesis from Unposed Imagescited by 36
  3. 2026NitroGen: An Open Foundation Model for Generalist Gaming Agentscited by 27
  4. 2025Epipolar Geometry Improves Video Generation Modelscited by 26
  5. 2024GSplatLoc: Grounding Keypoint Descriptors into 3D Gaussian Splatting for Improved Visual Localizationcited by 16
  6. 2025JamMa: Ultra-lightweight Local Feature Matching with Joint Mambacited by 14
  7. 2025From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splattingcited by 14
  8. 2024Learning to Make Keypoints Sub-Pixel Accuratecited by 13
  9. 2024ASIGN: An Anatomy-aware Spatial Imputation Graphic Network for 3D Spatial Transcriptomicscited by 13
  10. 2025RDD: Robust Feature Detector and Descriptor using Deformable Transformercited by 13
  11. 2025MATCHA:Towards Matching Anythingcited by 12
  12. 2025LiftFeat: 3D Geometry-Aware Local Feature Matchingcited by 12
38 more
  1. 2025OrthoLoC: UAV 6-DoF Localization and Calibration Using Orthographic Geodatacited by 12
  2. 20253R-GS: Best Practice in Optimizing Camera Poses Along with 3DGScited by 11
  3. 2026Articulated 3D Scene Graphs for Open-World Mobile Manipulationcited by 11
  4. 2025Alligat0R: Pre-Training Through Co-Visibility Segmentation for Relative Camera Pose Regressioncited by 10
  5. 2025Multi-Resolution SAR and Optical Remote Sensing Image Registration Methods: A Review, Datasets, and Future Perspectivescited by 9
  6. 2025RUBIK: A Structured Benchmark for Image Matching across Geometric Challengescited by 8
  7. 2025One-Shot Dual-Arm Imitation Learningcited by 8
  8. 2025VGGT-X: When VGGT Meets Dense Novel View Synthesiscited by 8
  9. 2025EDM: Efficient Deep Feature Matchingcited by 7
  10. 2025Deep Learning Reforms Image Matching: A Survey and Outlookcited by 7
  11. 2024Affine steerers for structured keypoint descriptioncited by 6
  12. 2025MIFNet: Learning Modality-Invariant Features for Generalizable Multimodal Image Matchingcited by 6
  13. 2024Leveraging Semantic Cues from Foundation Vision Models for Enhanced Local Feature Correspondencecited by 5
  14. 2025MapGlue: Multimodal Remote Sensing Image Matchingcited by 5
  15. 2025Splat and Replace: 3D Reconstruction with Repetitive Elementscited by 5
  16. 2025DINO-VO: A Feature-based Visual Odometry Leveraging a Visual Foundation Modelcited by 5
  17. 2024Mismatched: Evaluating the Limits of Image Matching Approaches and Benchmarkscited by 4
  18. 2024LiteVLoc: Map-Lite Visual Localization for Image Goal Navigationcited by 4
  19. 2025Learning Dense Feature Matching via Lifting Single 2D Image to 3D Spacecited by 4
  20. 2025pySLAM: An Open-Source, Modular, and Extensible Framework for SLAMcited by 3
  21. 2025To Match or Not to Match: Revisiting Image Matching for Reliable Visual Place Recognitioncited by 3
  22. 2025Unsupervised Deformable Image Registration with Structural Nonparametric Smoothingcited by 3
  23. 2025FLOWING: Implicit Neural Flows for Structure-Preserving Morphingcited by 3
  24. 2025Changes in Real Time: Online Scene Change Detection with Multi-View Fusioncited by 3
  25. 2025On-the-fly Large-scale 3D Reconstruction from Multi-Camera Rigscited by 3
  26. 2026Second-order Gaussian directional derivative representations for image high-resolution corner detectioncited by 3
  27. 2026RaCo: Ranking and Covariance for Practical Learned Keypointscited by 3
  28. 2026CRFT: Consistent-Recurrent Feature Flow Transformer for Cross-Modal Image Registrationcited by 3
  29. 2024Scalable and Generalizable Correspondence Pruning via Geometry-Consistent Pre-trainingcited by 2
  30. 2024XPoint: A Self-Supervised Visual-State-Space based Architecture for Multispectral Image Registrationcited by 2
  31. 2025AlignDiff: Learning Physically-Grounded Camera Alignment via Diffusioncited by 2
  32. 2025Hierarchical Image Matching for UAV Absolute Visual Localization via Semantic and Structural Constraintscited by 2
  33. 2025BleedOrigin: Dynamic Bleeding Source Localization in Endoscopic Submucosal Dissection via Dual-Stage Detection and Trackingcited by 2
  34. 2025LightGlueStick: a Fast and Robust Glue for Joint Point-Line Matchingcited by 2
  35. 2026A Streamlined Attention-Based Network for Descriptor Extractioncited by 2
  36. 2026Are Pretrained Image Matchers Good Enough for SAR-Optical Satellite Registration?cited by 2
  37. 2026Region Matters: Efficient and Reliable Region-Aware Visual Place Recognitioncited by 2
  38. 2026Immediate 3D Gaussian Splat Reconstruction of Unordered Input with Global Consistencycited by 2

50 most cited of 113 All 113 (JSON)