AI Paper Atlas

Research map / Video / Video-language understanding

MomentDiff: Generative Video Moment Retrieval from Random to Real

Pandeng Li, Chen-Wei Xie, Hongtao Xie, Liming Zhao, Lei Zhang, Yun Zheng, Deli Zhao, Yongdong Zhang

NeurIPS 2023 · Video › Video-language understanding · approach: Video moment retrieval · 23 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 random-span diffusion denoising instead of learned proposalsMomentDiff · Charades-STA-Len and Charades-STA-MomR1@0.5: 34.39 -> 38.32; 22.16 -> 33.59Improved
Replace FC span embedding with ROI pooling (ablation)MomentDiff · Charades-STAR1@0.5: 55.57 -> 50.21Hurt
Set diffusion scale to 1.0 instead of 2.0 (ablation)MomentDiff · Charades-STAR1@0.5: 55.57 -> 54.33Hurt
Set diffusion scale to 3.0 instead of 2.0 (ablation)MomentDiff · Charades-STAR1@0.5: 55.57 -> 50.12Hurt
Add the VMD denoiser and diffusion process (ablation)MomentDiff · Charades-STAR1@0.5: 46.69 -> 55.57Improved
Add explicit noise-intensity conditioning (ablation)MomentDiff · Charades-STAR1@0.5: 50.41 -> 55.57Improved
Add the VMR localization loss alongside similarity loss (ablation)MomentDiff · Charades-STAR1@0.5: 31.53 -> 55.57Improved
Add similarity loss alongside VMR loss (ablation)MomentDiff · Charades-STAR1@0.5: 40.94 -> 55.57Improved
Use 5 random spans rather than 1, 10, or 20 (ablation)MomentDiff · Charades-STAR1@0.5: 50.83; 55.57; 53.89; 50.42Mixed
Increase diffusion sampling steps (ablation)MomentDiff · Charades-STAR1@0.7: 29.78 -> 32.42Improved
Use DDIM instead of DDPM sampling (ablation)MomentDiff · Charades-STAMAP_avg: 32.48 -> 32.85Improved
Use a cosine rather than linear beta schedule (ablation)MomentDiff · Charades-STAR1@0.5: 54.76 -> 55.57Improved

Later papers that compared with this method

1 later accepted paper reports MomentDiff as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.

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