Object retrieval with large vocabularies and fast spatial matching

  title={Object retrieval with large vocabularies and fast spatial matching},
  author={James Philbin and Ondřej Chum and Michael Isard and Josef Sivic and Andrew Zisserman},
  journal={2007 IEEE Conference on Computer Vision and Pattern Recognition},
In this paper, we present a large-scale object retrieval system. [] Key Method To address this problem we compare different scalable methods for building a vocabulary and introduce a novel quantization method based on randomized trees which we show outperforms the current state-of-the-art on an extensive ground-truth. Our experiments show that the quantization has a major effect on retrieval quality. To further improve query performance, we add an efficient spatial verification stage to re-rank the results…

Figures and Tables from this paper

Total Recall: Automatic Query Expansion with a Generative Feature Model for Object Retrieval

This paper brings query expansion into the visual domain via two novel contributions: strong spatial constraints between the query image and each result allow us to accurately verify each return, suppressing the false positives which typically ruin text-based query expansion.

Lost in quantization: Improving particular object retrieval in large scale image databases

The state of the art in visual object retrieval from large databases is achieved by systems that are inspired by text retrieval. A key component of these approaches is that local regions of images

Boosting Object Retrieval With Group Queries

A discriminative criterion for improving result quality is proposed, which lends itself to the addition of extra query data, and it is shown that multiple query images can be combined to produce enhanced results.

Towards Large-scale Image Retrieval with a Disk-only Index

This paper argues that it is worth reconsidering image retrieval with local features because since then, large solid-state disks providing dramatically shorter access times, and more discriminative models enhancing the local features by encoding their spatial neighborhood using features from convolutional neural networks resulting in way fewer random read memory accesses are available.

All About VLAD

It is shown that a simple change to the normalization method significantly improves retrieval performance and vocabulary adaptation can substantially alleviate problems caused when images are added to the dataset after initial vocabulary learning.

Efficient and robust image ranking for object retrieval.

The definition of the object retrieval problem is revisited and a group-query method is proposed, in which the query is a collection of images depicting the same object instead of a single query image used in the traditional “query-byexample” methods.

Fast large-scale object retrieval with binary quantization

This paper proposes a feature quantization algorithm called binary quantization, which converts a scale-invariant feature transform (SIFT) feature into a descriptive and discriminative bit-vector, which allows itself to adapt to the classic inverted file structure for box indexing.

Neighborhood Matching for Image Retrieval

A geometric method known as neighborhood matching (NM), which revisits the keypoint matching process by considering a neighborhood around each keypoint and improves the efficiency of a geometric verification step in the image search system.

Dataset adaptation for the evaluation of small object retrieval in heavily cluttered background

  • D. Manger
  • Computer Science
    Commercial + Scientific Sensing and Imaging
  • 2018
This work proposes to adapt public datasets to generate novel evaluation setups yielding tasks that are closer to the problem of small object retrieval, and shows that the new evaluation setup allows focusing on specific characteristics such as the object size more easily during evaluation.

Improving image retrieval by introducing locality sensitive encoding to visual similarity measures

A probabilistic framework to model featureto-feature similarity for high-dimensional local features and a function to score the individual feature- to-feature contributions to an image-to-image similarity are presented, which consistently gives a significant boost to retrieval accuracy.



Scalable Recognition with a Vocabulary Tree

  • D. NistérHenrik Stewénius
  • Computer Science
    2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)
  • 2006
A recognition scheme that scales efficiently to a large number of objects and allows a larger and more discriminatory vocabulary to be used efficiently is presented, which it is shown experimentally leads to a dramatic improvement in retrieval quality.

Video Google: a text retrieval approach to object matching in videos

We describe an approach to object and scene retrieval which searches for and localizes all the occurrences of a user outlined object in a video. The object is represented by a set of viewpoint

Sub-linear Indexing for Large Scale Object Recognition

A method capable of recognising one of N objects in log(N) time, which preserves all the strengths of local affine region methods – robustness to background clutter, occlusion, and large changes of viewpoints.

Shape Quantization and Recognition with Randomized Trees

A new approach to shape recognition based on a virtually infinite family of binary features (queries) of the image data, designed to accommodate prior information about shape invariance and regularity, and a comparison with artificial neural networks methods is presented.

Multiple Object Class Detection with a Generative Model

The performance of the proposed multi-object class detection approach is competitive to state of the art approaches dedicated to a single object class recognition problem.

Simultaneous Object Recognition and Segmentation by Image Exploration

A novel Object Recognition approach which overcomes limitations in dealing with extensive clutter, dominant occlusion, large scale and viewpoint changes, and can extend any viewpoint invariant feature extractor.

Multi-tier architecture for Web search engines

A novel multitier architecture to tier documents in a search engine is described, which allows for increased performance while keeping the order of the results returned, and hence relevance, almost "untouched".

Scale & Affine Invariant Interest Point Detectors

A comparative evaluation of different detectors is presented and it is shown that the proposed approach for detecting interest points invariant to scale and affine transformations provides better results than existing methods.

Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

New results are derived on the minimum number of landmarks needed to obtain a solution, and algorithms are presented for computing these minimum-landmark solutions in closed form that provide the basis for an automatic system that can solve the Location Determination Problem under difficult viewing.

Web Search for a Planet: The Google Cluster Architecture

Googless architecture features clusters of more than 15,000 commodity-class PCs with fault tolerant software that achieves superior performance at a fraction of the cost of a system built from fewer, but more expensive, high-end servers.