LIFT: Learned Invariant Feature Transform

2016/03/30 by Kwang Moo Yi, Eduard Trulls, Yi, Kwang Moo +5 · 1 voice · 22 citations
Computer Science · Engineering · #Advanced Image and Video Retrieval Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Multimodal Machine Learning Applications #Robotics and Sensor-Based Localization #cs.CV

paper · pdf · doi:10.48550/arxiv.1603.09114

openalex publication_date 2016/03/30 · arxiv published 2016/03/30 · arxiv updated 2016/07/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We introduce a novel Deep Network architecture that implements the full feature point handling pipeline, that is, detection, orientation estimation, and feature description. While previous works have successfully tackled each one of these problems individually, we show how to learn to do all three in a unified manner while preserving end-to-end differentiability. We then demonstrate that our Deep pipeline outperforms state-of-the-art methods on a number of benchmark datasets, without the need of retraining.

Citations

• Distinctive Image Features from Scale-Invariant Keypoints

Cited by

• Match-Any-Events: Zero-Shot Motion-Robust Feature Matching Across Wide Baselines for Event Cameras
• MARVO: Marine-Adaptive Radiance-aware Visual Odometry
• Fin3R: Fine-tuning Feed-forward 3D Reconstruction Models via Monocular Knowledge Distillation
• RaCo: Ranking and Covariance for Practical Learned Keypoints
• LightGlueStick: a Fast and Robust Glue for Joint Point-Line Matching
• DeepDetect: Learning All-in-One Dense Keypoints
• SaLon3R: Structure-aware Long-term Generalizable 3D Reconstruction from Unposed Images
• Guided Image Feature Matching using Feature Spatial Order
• HART: Human Aligned Reconstruction Transformer
• From Fields to Splats: A Cross-Domain Survey of Real-Time Neural Scene Representations
• SAIL-Recon: Large SfM by Augmenting Scene Regression with Localization
• Dual-Mode Deep Anomaly Detection for Medical Manufacturing: Structural Similarity and Feature Distance
• Human Pose Regression by Combining Indirect Part Detection and\n Contextual Information
• RegNet: Learning the Optimization of Direct Image-to-Image Pose Registration
• Large-Scale Image Retrieval with Attentive Deep Local Features
• G-CUT3R: Guided 3D Reconstruction with Camera and Depth Prior Integration
• A Performance Evaluation of Local Features for Image Based 3D Reconstruction
• Refining Gaussian Splatting: A Volumetric Densification Approach
• DOMR: Establishing Cross-View Segmentation via Dense Object Matching
• Learning Spread-out Local Feature Descriptors
• GeoMoE: Divide-and-Conquer Motion Field Modeling with Mixture-of-Experts for Two-View Geometry
• Modality-Aware Feature Matching: A Comprehensive Review of Single- and Cross-Modality Techniques
• Estimating 2D Camera Motion with Hybrid Motion Basis

Discussions

• LIFT: Learned Invariant Feature Transform [hn, 3 points, 0 comments]

Related

• Geometry-aware Deep Transform
• Deep Feature Surgery: Towards Accurate and Efficient Multi-Exit Networks
• S-TREK: Sequential Translation and Rotation Equivariant Keypoints for local feature extraction
• Feature Graph Architectures
• Feature Network Methods in Machine Learning and Applications