AI Paper Atlas

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

CAT: Enhancing Multimodal Large Language Model to Answer Questions in Dynamic Audio-Visual Scenarios

Qilang Ye, Zitong Yu, Rui Shao, Xinyu Xie, Philip Torr, Xiaochun Cao

ECCV 2024 · Vision-language and multimodal › Visual question answering · approach: Multimodal question answering datasets · 37 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 question-related clue tokens to video and audio tokens (ablation)CAT · Video-ChatGPT benchmarkCorrectness of Information: 40.6 -> 61.6Improved
add audio tokens to video tokens (ablation)CAT · Video-ChatGPT benchmarkCorrectness of Information: 38.6 -> 40.6Improved
add visual query h_v in clue aggregator (ablation)CAT clue aggregator · AVinstruct test setBLEU-4: 10.0 -> 30.8Improved
add audio query h_a in clue aggregator (ablation)CAT clue aggregator · AVinstruct test setBLEU-4: 18.8 -> 30.8Improved
add reverse-order block B2 after B1 (ablation)CAT clue aggregator · AVinstruct test setBLEU-4: 25.9 -> 30.8Improved
add ADPO retraining (ablation)CAT · AVSDCIDEr: 74.8 -> 79.0Improved
add ADPO retraining (ablation)Video-LLaMA · AVSDCIDEr: 63.7 -> 69.2Improved
increase number of visual tokens (ablation)CAT · Video-ChatGPT benchmarktemporal/consistency/detail accuracyHurt
use 48 learnable queries in Q-former for clue tokens (ablation)CAT clue aggregator · Video-ChatGPT benchmarkaccuracyImproved
use CAT-7B instead of ChatBridge-13B in zero-shot Music-AVQAaudio-visual MLLM · Music-AVQA (zero-shot)accuracy: 47.6 -> 48.6Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 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