AI Paper Atlas

Research map / Video / Video generation

V-Trans4Style: Visual Transition Recommendation for Video Production Style Adaptation

Pooja Guhan, Tsung-Wei Huang, Guan-Ming Su, Subhadra Gopalakrishnan, Dinesh Manocha

ECCV 2024 · Video › Video generation · approach: Compositional text-to-video diffusion · 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
replace AutoTransition with the transformer encoder-decoder with masked triplet losstransition recommendation · AutoTransition++ (pre-trained transition embedding)Recall@5: 44.58 -> 56.06Improved
use pre-trained transition embeddings instead of random embeddings (ablation)V-Trans4Style (E+D) · AutoTransition++Recall@5: 47.39 -> 56.06Improved
use MLN embeddings instead of AutoTransition's embeddings (ablation)V-Trans4Style (E+D) · AutoTransition++Recall@1 / Recall@5: Recall@1 23.24 -> 22.59; Recall@5 56.06 -> 61.03Mixed
add masked triplet loss to classification loss (ablation)V-Trans4Style (E+D) · AutoTransition++Recall@1: 11.63 -> 22.59Improved
add classification loss to masked triplet loss (ablation)V-Trans4Style (E+D) · AutoTransition++Mean rank: 8.57 -> 5.90Improved
triplet loss alone versus classification loss alone (ablation)V-Trans4Style (E+D) · AutoTransition++Recall@5: 69.00 -> 45.79Hurt
add style conditioning module without RRT (ablation)V-Trans4Style (E&D) · Style cosine similarity, Vlogcosine similarity: -0.01 -> 0.018Improved
add RRT fine-tuning of the transition sequence to SCM (ablation)V-Trans4Style SCM · Style cosine similarity, Photoscosine similarity: 0.0211 -> 0.7286Improved

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