AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Joint Metrics Matter: A Better Standard for Trajectory Forecasting

Erica Weng, Hana Hoshino, Deva Ramanan, Kris Kitani

ICCV 2023 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Interactive multi-agent trajectory prediction · 25 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
add joint (min-over-samples of agent-averaged) reconstruction loss to AgentFormer trainingAgentFormer · ETH/UCY avgmin JADE (m): 0.384 -> 0.357Improved
add joint loss to AgentFormerAgentFormer · ETH/UCY avgCR_mean: 0.063 -> 0.054Improved
add joint loss to AgentFormerAgentFormer · ETH/UCY avgmin ADE (m): 0.233 -> 0.245Hurt
add joint loss to View VerticallyView Vertically · ETH/UCY avgmin JADE (m): 0.408 -> 0.392Improved
add joint loss to View VerticallyView Vertically · ETH/UCY avgCR_mean: 0.077 -> 0.088Hurt
add joint loss on dense scenes versus sparse scenesJoint AgentFormer / Joint View Vertically · ETH/UCY scenesJADE: 18% and 10% better on zara2; smaller margin on eth, hotel, SDDMixed
use joint loss only in both AgentFormer training steps instead of marginal+joint in step 1 (ablation)Joint AgentFormer · ETH/UCY avgmin JADE (m): 0.358 -> 0.386Hurt
add joint term in step 1 of AgentFormer training (ablation)Joint AgentFormer · ETH/UCY avgmin JADE (m): 0.365 -> 0.358Improved
add joint loss in the first training step (ablation)Joint AgentFormer · ETH/UCY avgCR_mean: 0.064 -> 0.053Improved
joint-only loss in both steps (ablation)AgentFormer · ETH/UCY avgmin ADE (m): 0.233 -> 0.293Hurt
joint loss reduces collisions by interaction categoryJoint AgentFormer · ETH/UCY interaction categoriesCR_mean: 0.103 -> 0.084 (group)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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