Avoiding Confusing Features in Place Recognition

By Jan Knopp, Josef Sivic, Tomás Pajdla. 2010

249 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

CPlaNet: Enhancing Image Geolocalization by Combinatorial Partitioning of Maps (2018, background): “While global features have the beneﬁt to retrieve diverse natural scene images based on ambient information, local image features yield higher precision in retrieving structured objects such as buildings and are thus more frequently used [8,9,10,11,12,13,14,15,16].”
PlaNet - Photo Geolocation with Convolutional Neural Networks (2016, background): “Most local feature based approaches therefore focus on localization within cities, either based on photos from photo sharing websites [5, 36] or street view [2, 7, 26, 27, 43, 51, 52].”
What makes Paris look like Paris? (2012, neutral): “However, such systems require signiﬁcant manual e ↵ ort from an expert to specify the appropriate elements and rules for each architectural style.”
Google Landmarks Dataset v2 – A Large-Scale Benchmark for Instance-Level Recognition and Retrieval (2020, background): “Other datasets also focus on imagery from a single city: Rome 16 k [1]; Geotagged Streetview Images [32] containing 17 k images from Paris; San Francisco Landmarks [14] containing 1 .”
24/7 Place Recognition by View Synthesis (2018, unknown): “The large-scale place recognition is often formulated as a variation of image retrieval [22, 33] where the query photograph is localized by matching it to a large database of geo-tagged images such as Google street-view [6, 8, 10, 14, 24, 35, 36, 40, 44].”
Accurate Geo-Registration by Ground-to-Aerial Image Matching (2014, background): “Recently, researchers have looked into using other sources of geo-location proxies, for example, matching ground images to geo-located Google Street View photos [8][27], or looking for GPS tagged images with similar appearance [11][15], matching to geolocated ground-level 3D points [17].”
Total recall II: Query expansion revisited (2011, background): “In the literature, the groups have been called coocsets [6], or confusers [13].”
Fast image-based localization using direct 2D-to-3D matching (2011, background): “In this case, it might be necessary to remove confusing descriptors [15] or to initially allow ambiguous matches which are resolved by RANSAC [12].”
Image Geo-Localization Based on MultipleNearest Neighbor Feature Matching UsingGeneralized Graphs (2014, methodology): “First, we extract local features (e.g., SIFT) from the query image and retrieve a number of nearest neighbors for each query feature from the reference data set.”
Graph-Based Discriminative Learning for Location Recognition (2013, methodology): “In this way, the image graph G gives us the supervision necessary to define positives and negatives for learning, just as geotags have provided a supervisory cue for discriminative location recognition in previous work [27, 14].”
Learning and Calibrating Per-Location Classifiers for Visual Place Recognition (2013, methodology): “Visual place recognition [7, 13, 27] is a challenging task as the query and database images may depict the same 3D structure (e.g. a building) from a different camera view-point, under different illumination, or the building can be partially occluded.”
Image Retrieval for Image-Based Localization Revisited (2012, background): “Schindler et al. select only informative features [29] to construct a vocabulary tree [22] while Knopp et al. remove confusing features [16].”
Efficient & Effective Prioritized Matching for Large-Scale Image-Based Localization (2017, background): “Over the years, various improvements have been made to handle confusing [15] and repetitive [16] structures, scale to larger scenes [14], [17], or use the original feature descriptors for retrieval [18], [19].”
Three things everyone should know to improve object retrieval (2012, methodology): “Recent work [5, 13] has proposed methods for identifying words that are confusing for a given query image, and thus should not be used.”
Wide-Area Image Geolocalization with Aerial Reference Imagery (2015, background): “In recent years, a plethora of methods for automatic image geolocalization have been introduced [1, 8, 11, 19, 23, 45].”
Visual Place Recognition with Repetitive Structures (2015, methodology): “While repeated structures often occur in man-made environments – examples include building facades, fences, or road markings – they are usually treated as nuisance and downweighted at the indexing stage [13, 18, 36, 39].”
Leveraging Structure from Motion to Learn Discriminative Codebooks for Scalable Landmark Classification (2013, background): “Other aggregation strategies such as learning feature weights have also been proposed to handle background clutter [1, 14] Although unsupervised codebooks continue to be popular, discriminative codebooks are known to be superior [32] as they aim to encode the differences between categories [24].”