AI Paper Atlas

Research map / Video / Action recognition

Transferring Knowledge From Text to Video: Zero-Shot Anticipation for Procedural Actions

Fadime Şener, Rishabh Saraf, Angela Yao

TPAMI 2022 · Video › Action recognition · approach: Disentangled action recognition · 24 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
pre-train the text model on more of Recipe1M (25% to 100%) (ablation)ours text · Tasty Videos V1BLEU4: 2.03 -> 3.30Improved
pre-train on 50% instead of 0% of Recipe1M (ablation)ours text · Tasty Videos V1BLEU4: 0.24 -> 2.42Improved
add recipe loss L_r to decoder loss L_d (ablation)recipe network · Recipe1M test, next stepingredient recall: 32.3 -> 33.5Improved
add recipe loss L_r to decoder loss L_d (ablation)recipe network · Recipe1M test, next+3 stepBLEU4: 1.3 -> 0.4Hurt
use beam search (k=5) instead of greedy decoding (ablation)recipe network · Recipe1M test, next stepingredient recall: 32.3 -> 33.1Improved
use ground-truth segments instead of fixed windows (ablation)ours visual · Tasty Videos V1BLEU4: 1.23 -> 1.48Improved
remove the ingredient list input (ablation)ours text · Tasty Videos V1BLEU4: 3.30 -> 0.92Hurt
add accompanying text with video (joint training) (ablation)ours visual (window) · Tasty Videos V1BLEU4: 1.23 -> 2.33Improved
use segment proposals instead of fixed windows (ablation)ours video · Tasty Videos V2BLEU4: 1.42 -> 1.63Improved
use Recipe1M pre-training in supervised setting (ablation)sup. visual (GT) · YouCookII 4-foldBLEU4: 0.15 -> 1.21Improved
use zero-shot instead of supervised training on same dishes (ablation)visual (GT) · YouCookII 4-foldBLEU4: 1.21 -> 0.84Hurt
anticipate next step versus caption current step with S2VTS2VT · Tasty Videos V1BLEU4: 1.10 -> 0.26Hurt

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