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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Recently, the methods based on Convolutional Neural Networks (CNNs) havegained popularity in the field of visual place recognition (VPR). Inparticular, the features from the middle layers of CNNs are more robust todrastic appearance changes than handcrafted features and high-layer features.Unfortunately, the holistic mid-layer features lack robustness to largeviewpoint changes. Here we split the holistic mid-layer features into localfeatures, and propose an adaptive dynamic time warping (DTW) algorithm to alignlocal features from the spatial domain while measuring the distance between twoimages. This realizes viewpoint-invariant and condition-invariant placerecognition. Meanwhile, a local matching DTW (LM-DTW) algorithm is applied toperform image sequence matching based on temporal alignment, which achievesfurther improvements and ensures linear time complexity. We perform extensiveexperiments on five representative VPR datasets. The results show that theproposed method significantly improves the CNN-based methods. Moreover, ourmethod outperforms several state-of-the-art methods while maintaining goodrun-time performance. This work provides a novel way to boost the performanceof CNN methods without any re-training for VPR. The code is available athttps://github.com/Lu-Feng/STA-VPR.

STA-VPR: Spatio-Temporal Alignment for Visual Place Recognition