AI Paper Atlas

Research map / Video / Video generation

Early Anticipation of Driving Maneuvers

Abdul Wasi Lone, Shankar Gangisetty, Shyam Nandan Rai, C. V. Jawahar

ECCV 2024 · Video › Video generation · approach: World-model video generation · 2 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 gaze modality to multi-view input (ablation)M2MVT · DAADaccuracy: 63.28 -> 65.59Improved
add front view (ablation)M2MVT · DAADaccuracy: 60.53 -> 65.59Improved
add in-cabin driver view (ablation)M2MVT · DAADaccuracy: 63.82 -> 65.59Improved
increase episodic memory tokens from 0 to 12 (ablation)M2MVT · DAADaccuracy: 63.46 -> 65.59Improved
increase memory tokens from 12 to 16 (ablation)M2MVT · DAADaccuracy: 65.59 -> 65.35Hurt
replace ViT encoder with M2MVT hybrid-fusion multiscale encoderCEMFormer · DAAD (all views + gaze)accuracy: 60.74 -> 65.59Improved
replace MViTv2 encoder with M2MVT fusion designMViTv2 · DAAD (all views + gaze)accuracy: 62.78 -> 65.59Improved
train and test on longer videos including before-maneuver (BM) context rather than only maneuver (DM) clipsM2MVT · DAADanticipation accuracy before onsetImproved
move from short DIP datasets to longer DAAD videosexisting DIP baselines · DAAD vs Brain4Carsaccuracy: 83.37 -> 58.33 (CEMFormer, in and out-cabin)Hurt

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