English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

With the recent success of visual features from deep convolutional neuralnetworks (DCNN) in visual robot self-localization, it has become important andpractical to address more general self-localization scenarios. In this paper,we address the scenario of self-localization from images with small overlap. Weexplicitly introduce a localization difficulty index as a decreasing functionof view overlap between query and relevant database images and investigateperformance versus difficulty for challenging cross-view self-localizationtasks. We then reformulate the self-localization as a scalablebag-of-visual-features (BoVF) scene retrieval and present an efficient solutioncalled PCA-NBNN, aiming to facilitate fast and yet discriminativecorrespondence between partially overlapping images. The proposed approachadopts recent findings in discriminativity preserving encoding of DCNN featuresusing principal component analysis (PCA) and cross-domain scene matching usingnaive Bayes nearest neighbor distance metric (NBNN). We experimentallydemonstrate that the proposed PCA-NBNN framework frequently achieves comparableresults to previous DCNN features and that the BoVF model is significantly moreefficient. We further address an important alternative scenario of"self-localization from images with NO overlap" and report the result.

Self-localization from Images with Small Overlap