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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Multi-modality fusion and multi-task learning are becoming trendy in 3Dautonomous driving scenario, considering robust prediction and computationbudget. However, naively extending the existing framework to the domain ofmulti-modality multi-task learning remains ineffective and even poisonous dueto the notorious modality bias and task conflict. Previous works manuallycoordinate the learning framework with empirical knowledge, which may lead tosub-optima. To mitigate the issue, we propose a novel yet simple multi-levelgradient calibration learning framework across tasks and modalities duringoptimization. Specifically, the gradients, produced by the task heads and usedto update the shared backbone, will be calibrated at the backbone's last layerto alleviate the task conflict. Before the calibrated gradients are furtherpropagated to the modality branches of the backbone, their magnitudes will becalibrated again to the same level, ensuring the downstream tasks pay balancedattention to different modalities. Experiments on large-scale benchmarknuScenes demonstrate the effectiveness of the proposed method, eg, an absolute14.4% mIoU improvement on map segmentation and 1.4% mAP improvement on 3Ddetection, advancing the application of 3D autonomous driving in the domain ofmulti-modality fusion and multi-task learning. We also discuss the linksbetween modalities and tasks.

FULLER: Unified Multi-modality Multi-task 3D Perception via Multi-level Gradient Calibration