RTAformer tackles 'relational semantic inertia' in video scene graph generation
Jaeyeon Heo · hf · 2026-10-23
Researchers at Sejong University propose RTAformer, a one-stage frame-work transformer for video scene graph generation. It identifies a failure mode called relational semantic inertia, where models keep stale relation predictions after relationships change. The method separately models entity persistence vs. relation change, injects calibrated differences between adjacent-frame relation queries to guide predicate updates, and uses cross-frame instance alignment during training to reduce representational drift.
More from Research
- CRYSTALMINER: an OSI hackathon project teaching crystallography via 23 Minecraft-like symmetry worlds with an AI tutor — CatAstro_Piyush · 2026-10-23
- Lucas Beyer counters Nesterov's generalization pessimism with the transformer inductive-bias argument — giffmana · 2026-10-23
- First-ever in-body prime editing treatment injected into a human — basedjensen · 2026-10-23
- A classic world model (s, a → s′) acting as a neural MuJoCo simulator — yshan2u · 2026-10-23
- Gen-HumanWholeBody: open 600-hour whole-body manipulation dataset for robots — AdinaYakup · 2026-10-23
- DynaVelo, a neural ODE model of single-cell dynamics, published in Nature Biotechnology — anshulkundaje · 2026-10-23