AI Paper Atlas

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

Segment-Then-Rank: Non-Factoid Question Answering on Instructional Videos

Kyungjae Lee, Nan Duan, Lei Ji, Jason Li, Seung-won Hwang

AAAI 2020 · Vision-language and multimodal › Visual question answering · approach: Multimodal question answering datasets · 1 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 Ranker on top of Segmenter (two-stage) (ablation)Segmenter · Instructional video non-factoid QA test setF1: 65.71 -> 70.31Improved
Add object features to the encoder (ablation)Segmenter · Instructional video non-factoid QA test setrecall: 64.45 -> 67.21Improved
Add object features to the encoder (ablation)Segmenter + Ranker · Instructional video non-factoid QA test setrecall: 71.33 -> 74.75Improved
Add object features to the encoder (ablation)Segmenter + Ranker · Instructional video non-factoid QA test setF1: 68.53 -> 70.31Improved
Add self-attention and LSTM hierarchical modeling (ablation)Segmenter · Instructional video non-factoid QA test setF1: 51.31 -> 65.71Improved
Add LSTM on top of self-attention-free Segmenter (ablation)Segmenter · Instructional video non-factoid QA test setF1: 51.31 -> 64.08Improved
Add self-attention to Segmenter (ablation)Segmenter · Instructional video non-factoid QA test setF1: 64.08 -> 65.71Improved
Use interaction-based features instead of representation-based (ablation)Segmenter + Ranker · Instructional video non-factoid QA test setF1: 67.57 -> 68.57Improved
Concatenate representation and interaction features (ablation)Segmenter + Ranker · Instructional video non-factoid QA test setF1: 68.57 -> 68.74Improved
Use length-adaptive gate instead of concatenation (ablation)Segmenter + Ranker · Instructional video non-factoid QA test setF1: 68.74 -> 70.31Improved
Use Segmenter+Ranker instead of Ranker over fixed windowsRanker · Instructional video non-factoid QA test setF1: 53.97 -> 70.31Improved
Use BERT encoder with Segmenter instead of QANet encoderBase-III · Instructional video non-factoid QA test setF1: 61.57 -> 65.71Improved

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