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

We propose Compact and Swift Segmenting 3D Gaussians(CoSSegGaussians), amethod for compact 3D-consistent scene segmentation at fast rendering speedwith only RGB images input. Previous NeRF-based segmentation methods haverelied on time-consuming neural scene optimization. While recent 3D GaussianSplatting has notably improved speed, existing Gaussian-based segmentationmethods struggle to produce compact masks, especially in zero-shotsegmentation. This issue probably stems from their straightforward assignmentof learnable parameters to each Gaussian, resulting in a lack of robustnessagainst cross-view inconsistent 2D machine-generated labels. Our method aims toaddress this problem by employing Dual Feature Fusion Network as Gaussians'segmentation field. Specifically, we first optimize 3D Gaussians under RGBsupervision. After Gaussian Locating, DINO features extracted from images areapplied through explicit unprojection, which are further incorporated withspatial features from the efficient point cloud processing network. Featureaggregation is utilized to fuse them in a global-to-local strategy for compactsegmentation features. Experimental results show that our model outperformsbaselines on both semantic and panoptic zero-shot segmentation task, meanwhileconsumes less than 10\% inference time compared to NeRF-based methods. Code andmore results will be available at https://David-Dou.github.io/CoSSegGaussians.

CoSSegGaussians: Compact and Swift Scene Segmenting 3D Gaussians with Dual Feature Fusion