TILDE: A Temporally Invariant Learned DEtector

Abstract

We introduce a learning-based approach to detect repeatable keypoints under drastic imaging changes of weather and lighting conditions to which state-of-the-art keypoint detectors are surprisingly sensitive. We first identify good keypoint candidates in multiple training images taken from the same viewpoint. We then train a regressor to predict a score map whose maxima are those points so that they can be found by simple non-maximum suppression. As there are no standard datasets to test the influence of these kinds of changes, we created our own, which we will make publicly available. We will show that our method significantly outperforms the state-of-the-art methods in such challenging conditions, while still achieving state-of-the-art performance on the untrained standard Oxford dataset.

Cited by in corpus (46)

• 3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration
• Image Matching across Wide Baselines: From Paper to Practice
• DISK: Learning local features with policy gradient
• Mini-Unmanned Aerial Vehicle-Based Remote Sensing: Techniques, Applications, and Prospects
• Keyframe-based monocular SLAM: design, survey, and future directions
• LIFT: Learned Invariant Feature Transform
• UnsuperPoint: End-to-end Unsupervised Interest Point Detector and Descriptor
• Single-View Place Recognition under Seasonal Changes
• Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filters
• Large-Scale Image Retrieval with Attentive Deep Local Features
• FFD: Fast Feature Detector
• HarrisZ${}^{+}$: Harris Corner Selection for Next-Gen Image Matching Pipelines
• GCNv2: Efficient Correspondence Prediction for Real-Time SLAM
• Probabilistic Spatial Distribution Prior Based Attentional Keypoints Matching Network
• From handcrafted to deep local features
• Image Stylization for Robust Features
• A Comparison of CNN and Classic Features for Image Retrieval
• Local Feature Detectors, Descriptors, and Image Representations: A Survey
• AstroVision: Towards Autonomous Feature Detection and Description for Missions to Small Bodies Using Deep Learning
• Leveraging Outdoor Webcams for Local Descriptor Learning
• Improving the HardNet Descriptor
• Illumination-insensitive Binary Descriptor for Visual Measurement Based on Local Inter-patch Invariance
• Discovering Visual Patterns in Art Collections with Spatially-consistent Feature Learning
• A Performance Evaluation of Local Features for Image Based 3D Reconstruction
• RF-Net: An End-to-End Image Matching Network based on Receptive Field
• Optimizing Through Learned Errors for Accurate Sports Field Registration
• D2D: Keypoint Extraction with Describe to Detect Approach
• Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
• DiscoBox: Weakly Supervised Instance Segmentation and Semantic Correspondence from Box Supervision
• NeSS-ST: Detecting Good and Stable Keypoints with a Neural Stability Score and the Shi-Tomasi Detector
• SCK: A sparse coding based key-point detector
• Neural Outlier Rejection for Self-Supervised Keypoint Learning
• Quad-networks: unsupervised learning to rank for interest point detection
• Learning to Assign Orientations to Feature Points
• Soft Expectation and Deep Maximization for Image Feature Detection
• Aligning Across Large Gaps in Time
• Unsupervised Metric Relocalization Using Transform Consistency Loss
• DR-KFS: A Differentiable Visual Similarity Metric for 3D Shape Reconstruction
• IF-Net: An Illumination-invariant Feature Network
• DynaMiTe: A Dynamic Local Motion Model with Temporal Constraints for Robust Real-Time Feature Matching
• A Scale and Rotational Invariant Key-point Detector based on Sparse Coding
• Unsupervised Learning Framework of Interest Point Via Properties Optimization
• Learning to Predict Repeatability of Interest Points
• Correspondence-Free Pose Estimation with Patterns: A Unified Approach for Multi-Dimensional Vision
• DCI: Discriminative and Contrast Invertible Descriptor
• Learning Local Feature Descriptor with Motion Attribute for Vision-based Localization