An accurate retrieval through R-MAC+ descriptors for landmark recognition

Magliani, Federico; Prati, Andrea

doi:10.1145/3243394.3243686

Cited by 16 publications

(17 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This process is known as "transfer learning" and consists in tuning the parameters trained in one feature space in order to work in another feature space [21]. Some methods that use transfer learning are: Spatial pooling [25], MOP-CNN [7], Neural codes [3], Ng et al [34], CCS [33], OC [26], R-MAC [31], Gordo et al [8] and Magliani et al [18]. Also, fine-tuning global descriptors [9] on a similar image dataset, allows to highly improve accuracy results, but with an extra time effort due to the training phase on the new dataset.…”

Section: Related Workmentioning

confidence: 99%

A Dense-Depth Representation for VLAD Descriptors in Content-Based Image Retrieval

Magliani

Fontanini

Prati

2018

Advances in Visual Computing

Self Cite

View full text Add to dashboard Cite

The recent advances brought by deep learning allowed to improve the performance in image retrieval tasks. Through the many convolutional layers, available in a Convolutional Neural Network (CNN), it is possible to obtain a hierarchy of features from the evaluated image. At every step, the patches extracted are smaller than the previous levels and more representative. Following this idea, this paper introduces a new detector applied on the feature maps extracted from pre-trained CNN. Specifically, this approach lets to increase the number of features in order to increase the performance of the aggregation algorithms like the most famous and used VLAD embedding. The proposed approach is tested on different public datasets: Holidays, Oxford5k, Paris6k and UKB.

show abstract

Section: Related Workmentioning

confidence: 99%

A Dense-Depth Representation for VLAD Descriptors in Content-Based Image Retrieval

Magliani

Fontanini

Prati

2018

Advances in Visual Computing

Self Cite

View full text Add to dashboard Cite

show abstract

“…[53] proposes R-mac, also a pooling strategy, that was later trained end-to-end in [17] using the Siamese architecture. [28] improves upon R-mac [53] specifically for landmark recognition. In the same spirit, [46] uses R-mac in order to train CNN feature predictors for image retrieval.…”

Section: Related Workmentioning

confidence: 99%

“…Traditionally, vision-based localization [59] is tackled with structure-based methods, such as Structurefrom-Motion (SfM) [20,24,44,42,62,9] and Simultaneous Localization and Mapping (SLAM) [33,11,7,5,16], or with retrieval-based approaches [23,49,4,3,15,28,43,19]. Structure based methods usually focus on accurate relative pose estimation, while retrieval-based approaches prioritize absolute re-localization.…”

Section: Introductionmentioning

confidence: 99%

Learning Condition Invariant Features for Retrieval-Based Localization from 1M Images

Thoma,

Paudel,

Chhatkuli

et al. 2020

Preprint

View full text Add to dashboard Cite

Image features for retrieval-based localization must be invariant to dynamic objects (e.g. cars) as well as seasonal and daytime changes. Such invariances are, up to some extent, learnable with existing methods using triplet-like losses, given a large number of diverse training images. However, due to the high algorithmic training complexity, there exists insufficient comparison between different loss functions on large datasets. In this paper, we train and evaluate several localization methods on three different benchmark datasets, including Oxford RobotCar with over one million images. This large scale evaluation yields valuable insights into the generalizability and performance of retrieval-based localization. Based on our findings, we develop a novel method for learning more accurate and better generalizing localization features. It consists of two main contributions: (i) a feature volume-based loss function, and (ii) hard positive and pairwise negative mining. On the challenging Oxford RobotCar night condition, our method outperforms the well-known triplet loss by 24.4% in localization accuracy within 5m.

show abstract

“…Additionally, several papers have considered the application of embedding strategies on the feature maps extracted from the CNN (the most frequently used being, at present, VGG16 and ResNet101) in order to create a global descriptor. In fact, recently, several strategies have been proposed that are aimed at making the descriptors less sensitive to scale changes, rotation, occlusions, and so on [ 3 , 4 , 5 , 6 ].…”

Section: Introductionmentioning

confidence: 99%

Diffusion Parameters Analysis in a Content-Based Image Retrieval Task for Mobile Vision

Magliani

Sani

Cagnoni

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

Most recent computer vision tasks take into account the distribution of image features to obtain more powerful models and better performance. One of the most commonly used techniques to this purpose is the diffusion algorithm, which fuses manifold data and k-Nearest Neighbors (kNN) graphs. In this paper, we describe how we optimized diffusion in an image retrieval task aimed at mobile vision applications, in order to obtain a good trade-off between computation load and performance. From a computational efficiency viewpoint, the high complexity of the exhaustive creation of a full kNN graph for a large database renders such a process unfeasible on mobile devices. From a retrieval performance viewpoint, the diffusion parameters are strongly task-dependent and affect significantly the algorithm performance. In the method we describe herein, we tackle the first issue by using approximate algorithms in building the kNN tree. The main contribution of this work is the optimization of diffusion parameters using a genetic algorithm (GA), which allows us to guarantee high retrieval performance in spite of such a simplification. The results we have obtained confirm that the global search for the optimal diffusion parameters performed by a genetic algorithm is equivalent to a massive analysis of the diffusion parameter space for which an exhaustive search would be totally unfeasible. We show that even a grid search could often be less efficient (and effective) than the GA, i.e., that the genetic algorithm most often produces better diffusion settings when equal computing resources are available to the two approaches. Our method has been tested on several publicly-available datasets: Oxford5k, ROxford5k, Paris6k, RParis6k, and Oxford105k, and compared to other mainstream approaches.

show abstract

An accurate retrieval through R-MAC+ descriptors for landmark recognition

Cited by 16 publications

References 33 publications

A Dense-Depth Representation for VLAD Descriptors in Content-Based Image Retrieval

A Dense-Depth Representation for VLAD Descriptors in Content-Based Image Retrieval

Learning Condition Invariant Features for Retrieval-Based Localization from 1M Images

Diffusion Parameters Analysis in a Content-Based Image Retrieval Task for Mobile Vision

Contact Info

Product

Resources

About