AI Paper Atlas

Research map / Video / Video-language understanding

IPFormer: Instance Prompt-guided Transformer for Multi-modal Multi-shot Video Understanding

Yujia Liang, Jile Jiao, Xuetao Feng, Xinchen Liu, Kun Liu, Yuan Wang, Zixuan Ye, Hao Lu, Zhicheng Wang

AAAI 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 0 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 frame queries to Q-Former learnable queries (ablation)IPFormer compressor · ActivityNet-QAaccuracy: 40.9 -> 42.6Improved
add instance queries without clustering (ablation)IPFormer compressor · ActivityNet-QAaccuracy: 42.6 -> 43.6Improved
cluster instance tokens (ablation)IPFormer compressor · ActivityNet-QAaccuracy: 43.6 -> 45.5Improved
cluster instance tokens (ablation)IPFormer compressor · MultiClip-Benchaccuracy: 44.6 -> 46.2Improved
replace full projection by IPFormer compressorVideo-LLaVA · MultiClip-Benchaccuracy: 44.5 -> 46.2Improved
use plain Q-Former instead of full projection (ablation)Video-LLaVA · MultiClip-Benchaccuracy: 44.5 -> 42.3Hurt
use 2x2 average pooling instead of full projection (ablation)Video-LLaVA · MultiClip-Benchaccuracy: 44.5 -> 43.8Hurt
add IPFormer model design (no MultiClip data)Video-LLaVA · MultiClip-Benchmean accuracy: 44.5 -> 48.1Improved
add MultiClip training data (ablation)IPFormer · MultiClip-Benchmean accuracy: 48.1 -> 57.0Improved
add VideoChat2 instruction data (353k) (ablation)IPFormer · ActivityNet-QAaccuracy: 45.8 -> 48.5Improved
add MultiClip training data (ablation)PLLaVA · MultiClip-Benchaccuracy: 49.4 -> 57.6Improved
increase frames from 16 to 48IPFormer · MultiClip-Benchmean accuracy: 57.0 -> 58.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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