AI Paper Atlas

Research map / 3D vision / Human pose estimation

MonoLoco: Monocular 3D Pedestrian Localization and Uncertainty Estimation

Lorenzo Bertoni, Sven Kreiss, Alexandre Alahi

ICCV 2019 · 3D vision › Human pose estimation · approach: Transformer-based multi-view 3D pose · 114 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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
use Laplace loss instead of Gaussian loss (ablation)MonoLoco (PifPaf) · nuScenes teaser valALE (m): 1.97 -> 1.90Improved
use Laplace loss instead of L1 loss (ablation)MonoLoco (PifPaf) · nuScenes teaser valALE (m): 1.92 -> 1.90Improved
use Laplace loss instead of Gaussian loss (ablation)MonoLoco (Mask R-CNN) · nuScenes teaser valALE (m): 2.26 -> 2.12Improved
use Laplace loss instead of L1 loss (ablation)MonoLoco (Mask R-CNN) · nuScenes teaser valALE (m): 2.14 -> 2.12No clear effect
replace Mask R-CNN with PifPaf pose detector (ablation)MonoLoco (Laplace) · nuScenes teaser valALE (m): 2.12 -> 1.90Improved
replace geometric baseline with learned MonoLoco networkPifPaf poses · nuScenes teaser valALE (m): 2.05 -> 1.90Improved
replace geometric baseline with learned MonoLoco networkMask R-CNN poses · nuScenes teaser valALE (m): 3.73 -> 2.12Improved
train MonoLoco on nuScenes instead of KITTI (cross-dataset)MonoLoco · KITTI pedestriansALP <1m (%): 49.6 -> 51.7Improved
increase MC dropout probability from 0.05 to 0.4 (ablation)MonoLoco · KITTI valrecall (%): 82.8 -> 88.3Mixed
use MonoLoco instead of monocular MonoDepth + PifPafKITTI pedestrians · KITTI pedestriansALP <1m (%): 35.3 -> 51.7Improved
use monocular MonoLoco instead of stereo 3DOPKITTI pedestrians · KITTI pedestrians (Hard)ALE (m): 1.94 -> 1.17Improved
use monocular MonoLoco instead of stereo 3DOPKITTI pedestrians · KITTI pedestriansALP <0.5m (%): 41.4 -> 30.8Hurt

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