AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Zero-Shot Video Question Answering via Frozen Bidirectional Language Models

Antoine Yang, Antoine Miech, Josef Sivic, Ivan Laptev, Cordelia Schmid

NeurIPS 2022 · Vision-language and multimodal › Visual question answering · approach: Multimodal question answering datasets · 64 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 pretrained language model instead of random init (ablation)FrozenBiLM · MSVD-QA zero-shotaccuracy (%): 0.0 -> 31.9Improved
freeze language model weights (ablation)FrozenBiLM · MSVD-QA zero-shotaccuracy (%): 31.9 -> 32.2Improved
freeze language model weights (ablation)FrozenBiLM · How2QA zero-shotaccuracy (%): 45.7 -> 53.5Improved
freeze language model weights (ablation)FrozenBiLM · MSRVTT-QA zero-shotaccuracy (%): 17.6 -> 16.8Hurt
add adapters to frozen language model (ablation)FrozenBiLM · How2QA zero-shotaccuracy (%): 53.5 -> 58.4Improved
add adapters to frozen language model (ablation)FrozenBiLM · TVQA zero-shotaccuracy (%): 53.4 -> 59.2Improved
add adapters to frozen language model (ablation)FrozenBiLM · iVQA zero-shotaccuracy (%): 27.3 -> 26.8Hurt
add visual input (ablation)FrozenBiLM · iVQA zero-shotaccuracy (%): 13.2 -> 26.8Improved
add speech input to model with vision (ablation)FrozenBiLM · TVQA zero-shotaccuracy (%): 29.7 -> 59.2Improved
increase WebVid training data from 1K to 10M (ablation)FrozenBiLM · MSVD-QA zero-shotaccuracy (%): 13.6 -> 33.8Improved
replace autoregressive GPT-J-6B with bidirectional DeBERTa-V2-XLarge (ablation)FrozenBiLM · MSVD-QA zero-shotaccuracy (%): 26.7 -> 32.2Improved
increase bidirectional LM size from BERT-Base to DeBERTa-V2-XLarge (ablation)FrozenBiLM · MSVD-QA zero-shotaccuracy (%): 11.7 -> 32.2Improved

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