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

Anticipating motions of vehicles in a scene is an essential problem for safeautonomous driving systems. To this end, the comprehension of the scene'sinfrastructure is often the main clue for predicting future trajectories. Mostof the proposed approaches represent the scene with a rasterized format andsome of the more recent approaches leverage custom vectorized formats. Incontrast, we propose representing the scene's information by employing ScalableVector Graphics (SVG). SVG is a well-established format that matches theproblem of trajectory prediction better than rasterized formats while beingmore general than arbitrary vectorized formats. SVG has the potential toprovide the convenience and generality of raster-based solutions if coupledwith a powerful tool such as CNNs, for which we introduce SVG-Net. SVG-Net is aTransformer-based Neural Network that can effectively capture the scene'sinformation from SVG inputs. Thanks to the self-attention mechanism in itsTransformers, SVG-Net can also adequately apprehend relations amongst the sceneand the agents. We demonstrate SVG-Net's effectiveness by evaluating itsperformance on the publicly available Argoverse forecasting dataset. Finally,we illustrate how, by using SVG, one can benefit from datasets and advancementsin other research fronts that also utilize the same input format. Our code isavailable at https://vita-epfl.github.io/SVGNet/.

SVG-Net: An SVG-based Trajectory Prediction Model