AI Paper Atlas

Research map / Video / Video-language understanding

Learning Video Context as Interleaved Multimodal Sequences

Kevin Qinghong Lin, Pengchuan Zhang, Difei Gao, Xide Xia, Joya Chen, Ziteng Gao, Jinheng Xie, Xuhong Xiao, Mike Zheng Shou

ECCV 2024 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 5 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 subtitles to video input (ablation)MovieSeq · LVU relationaccuracy: 61.0 -> 75.6Improved
add subtitles to video input (ablation)MovieSeq · LVU sceneaccuracy: 76.8 -> 65.9Hurt
add subtitles to video input (ablation)MovieSeq · CMD video-text retrievalGeometric Mean: 9.2 -> 38.9Improved
add subtitles to video input (ablation)MovieSeq · TVC captioningCIDEr: 50.9 -> 63.5Improved
add character photos as context (ablation)MovieSeq · TVC captioningCIDEr: 63.5 -> 64.8No clear effect
add retrieved movie plot as context (ablation)MovieSeq · MovieQAaccuracy: 27.32 -> 66.41Improved
add movie plot context (ablation)MovieSeq · MADv2CIDEr: 11.4 -> 16.9Improved
add past clip narrations as context (ablation)MovieSeq · MADv2Rep@4: 1.33 -> 0.11Improved
character identification via Movienet-trained module instead of CLIP retrieval (ablation)MovieSeq · MADv2CIDEr: 19.7 -> 23.6Improved
multi-task co-training instead of individual training (ablation)MovieSeq · MAD/MovieNet/TVC dev setsaverage: 53.8 -> 55.2Improved
add subtitles on MovieQA video-only baseline (ablation)MovieSeq · MovieQAaccuracy: 27.32 -> 48.32Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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