AI Paper Atlas

Research map / Video / Video-language understanding

Prompt Switch: Efficient CLIP Adaptation for Text-Video Retrieval

Chaorui Deng, Qi Chen, Pengda Qin, Da Chen, Qi Wu

ICCV 2023 · Video › Video-language understanding · approach: Text-video retrieval · 56 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 Prompt Switch prompt cube to CLIP image encoder (ablation)CLIP ViT-B/32 · MSR-VTT 1K-AR@1: 43.1 -> 44.8Improved
add Prompt Aggregation on top of Prompt Switch (ablation)Prompt Switch · MSR-VTT 1K-AR@1: 44.8 -> 45.4Improved
add auxiliary captioning loss (ablation)Prompt Switch + Prompt Aggregation · MSR-VTT 1K-AR@1: 45.4 -> 46.1Improved
use Prompt Aggregation instead of Temporal Transformer (ablation)Prompt Switch · MSR-VTT 1K-AR@1: 44.2 -> 45.4Improved
use Token Shift for temporal modeling instead of Prompt Switch (ablation)CLIP · MSR-VTT 1K-AR@1: 45.4 -> 43.2Hurt
use full attention over all frame patches instead of Prompt Switch (ablation)CLIP · MSR-VTT 1K-AR@1: 45.4 -> 43.8Hurt
replace mean-pooling with X-Pool cross-modal fusion on top of Prompt Switch (ablation)Prompt Switch · MSR-VTT 1K-AR@1 (text-to-video): 46.1 -> 47.8Improved
use text-agnostic mean pooling instead of X-Pool fusion (ablation)Prompt Switch · MSR-VTT 1K-Aranking latency: 3.9s -> 8.1msImproved
randomly subsample k=3 of 6 prompt vectors during training (ablation)Prompt Switch · MSR-VTT 1K-Aretrieval performanceImproved
set captioning loss weight lambda to 0.5 instead of 0.3 (ablation)Prompt Switch · MSR-VTT 1K-Aretrieval performanceImproved
use mean-pool Prompt Switch instead of cross-modal methods on LSMDCPrompt Switch · LSMDCMeta Sum: 238.9 (X-Pool) vs 228.4Hurt

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