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

Prediction, decision-making, and motion planning are essential for autonomousdriving. In most contemporary works, they are considered as individual modulesor combined into a multi-task learning paradigm with a shared backbone butseparate task heads. However, we argue that they should be integrated into acomprehensive framework. Although several recent approaches follow this scheme,they suffer from complicated input representations and redundant frameworkdesigns. More importantly, they can not make long-term predictions about futuredriving scenarios. To address these issues, we rethink the necessity of eachmodule in an autonomous driving task and incorporate only the required modulesinto a minimalist autonomous driving framework. We propose BEVGPT, a generativepre-trained large model that integrates driving scenario prediction,decision-making, and motion planning. The model takes the bird's-eye-view (BEV)images as the only input source and makes driving decisions based onsurrounding traffic scenarios. To ensure driving trajectory feasibility andsmoothness, we develop an optimization-based motion planning method. Weinstantiate BEVGPT on Lyft Level 5 Dataset and use Woven Planet L5Kit forrealistic driving simulation. The effectiveness and robustness of the proposedframework are verified by the fact that it outperforms previous methods in 100%decision-making metrics and 66% motion planning metrics. Furthermore, theability of our framework to accurately generate BEV images over the long termis demonstrated through the task of driving scenario prediction. To the best ofour knowledge, this is the first generative pre-trained large model forautonomous driving prediction, decision-making, and motion planning with onlyBEV images as input.

BEVGPT: Generative Pre-trained Large Model for Autonomous Driving Prediction, Decision-Making, and Planning