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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Automated surgical gesture recognition is of great importance inrobot-assisted minimally invasive surgery. However, existing methods assumethat training and testing data are from the same domain, which suffers fromsevere performance degradation when a domain gap exists, such as the simulatorand real robot. In this paper, we propose a novel unsupervised domainadaptation framework which can simultaneously transfer multi-modalityknowledge, i.e., both kinematic and visual data, from simulator to real robot.It remedies the domain gap with enhanced transferable features by usingtemporal cues in videos, and inherent correlations in multi-modal towardsrecognizing gesture. Specifically, we first propose an MDO-K to alignkinematics, which exploits temporal continuity to transfer motion directionswith smaller gap rather than position values, relieving the adaptation burden.Moreover, we propose a KV-Relation-ATT to transfer the co-occurrence signals ofkinematics and vision. Such features attended by correlation similarity aremore informative for enhancing domain-invariance of the model. Two featurealignment strategies benefit the model mutually during the end-to-end learningprocess. We extensively evaluate our method for gesture recognition using DESKdataset with peg transfer procedure. Results show that our approach recoversthe performance with great improvement gains, up to 12.91% in ACC and 20.16% inF1score without using any annotations in real robot.

Domain Adaptive Robotic Gesture Recognition with Unsupervised Kinematic-Visual Data Alignment