AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Enhancing Vectorized Map Perception with Historical Rasterized Maps

Xiaoyu Zhang, Guangwei Liu, Zihao Liu, Ningyi Xu, Yunhui Liu, Ji Zhao

ECCV 2024 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 25 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add historical rasterized map (HRMapNet) to StreamMapNetStreamMapNet · nuScenes val (24 epochs)mAP: 60.4 -> 66.3Improved
add historical rasterized map (HRMapNet) to MapTRv2 with 110 epochsMapTRv2 · nuScenes val (110 epochs)mAP: 68.7 -> 73.6Improved
add historical rasterized mapStreamMapNet and MapTRv2 · Argoverse 2 valmAP: 61.5 -> 64.3 (StreamMapNet); 64.3 -> 68.3 (MapTRv2)Improved
add historical map without temporal fusion modules on location-disjoint splitsStreamMapNet · nuScenes and Argoverse 2 new splitsmAP: 33.9 -> 37.1 (nuScenes); 58.1 -> 61.5 (Argoverse 2)Improved
add BEV feature aggregation with local map (ablation)MapTRv2 · nuScenes val (24 epochs)mAP: 61.5 -> 64.6Improved
add query initialization with map prior embeddings (ablation)MapTRv2 + feature aggregation · nuScenes val (24 epochs)mAP: 64.6 -> 67.2Improved
downsample local map to 0.6 m before extracting prior embeddings (ablation)query initialization · nuScenes valmAP / GPU memory: mAP 66.3 -> 67.2; memory 64.94 GB -> 31.41 GBImproved
add Gaussian ego-pose noise at test time (ablation)HRMapNet · nuScenes valmAP: 67.2 -> 66.9 (0.1 m); 63.8 worst caseNo clear effect
start from validation-built or training-built global map instead of empty (ablation)HRMapNet · nuScenes valmAP: 67.2 -> 72.6 (validation map); 67.2 -> 83.7 (training map)Improved
use query initialization with fewer decoder layers (ablation)HRMapNet · nuScenes val (24 epochs)mAP: 64.6 (6 layers, no query init) -> 65.4 (4 layers, query init)Improved
add historical map under adverse conditions (night, rain)MapTRv2 · nuScenes night and rainymAP: night 39.6 -> 42.6; rainy 50.8 -> 62.6Improved
HRMapNet inference costMapTRv2 · nuScenes valFPS: 19.6 -> 17.0Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy