PlanRAG-Audio
PlanRAG-Audio: Planning and Retrieval Augmented Generation for Long-form Audio Understanding
Masao Someki, Chien-yu Huang, Siddhant Arora, Samuele Cornell, Markus Müller, Nathan Susanj, et al. · 2026-07-02
- Modality
- Audio & speech
- Data
- Real data
- Task · ground truth
- QA/Ret. · Ans.+Ev.+Time
- Scope
- Broader evaluation with memory demands
What it tests
- CECross-modal evidenceDirectly evaluated
- LELong-range evidenceDirectly evaluated
- SCState changesNot evaluated
- CFConflicting memoriesNot evaluated
- UQUnanswerable queriesNot evaluated
- MAMemory-grounded actionNot evaluated
- SPSpeaker trackingDirectly evaluated
- ACAcoustic-cue retentionDirectly evaluated
- RPRepair persistenceNot evaluated
Survey notes
Constructed long-audio QA evaluates perception, counting, temporal and compositional reasoning over recordings. The proposed retrieval agent is a method; its evaluation is not a distinct agent-memory retention or lifecycle benchmark.