StyleForge uses counterfactual reasoning to make room layouts more coherent
cn-scut · hf · 2026-08-04
StyleForge uses counterfactual reasoning to style indoor furniture layouts
The paper studies fixed-layout furniture styling: given a room layout, choose assets that fit the requested style without changing categories, positions, orientations, or scales.
Main idea
- A frozen multimodal LLM extracts structured style priors from the open-ended request and the layout.
- StyleForge maintains a learnable candidate distribution for each furniture slot.
- A dynamic hypergraph style field activates and weights layout-induced hyperedges to model higher-order dependencies among slots.
- Counterfactual style preference learning evaluates each candidate as a local substitution and scores contextual compatibility with Mahalanobis انرژی-like energies.
- Training alternates between the style field and candidate logits; at test time, only room-specific logits are updated.
Results
- Experiments on 3D-FRONT show better furniture retrieval and stronger scene-level style coherence than object-level and scene-level baselines.
More from Research
- GPT-5.6 Sol and Fable 5 reportedly proved an open best-of-n conjecture — abeirami · 2026-08-04
- A retina-inspired optical-flow sensor moves motion computation onto the chip — ssh4net · 2026-08-04
- Paper shows swap agnostic learning is equivalent to multicalibration and omniprediction — Sauers_ · 2026-08-04
- Critic says readers trusted the plots without reading the experiment section — suchenzang · 2026-08-04
- Embeddings can map huge historical news corpora, three papers show — leland_mcinnes · 2026-08-04
- Experiment finds the method hurts in mature action-policy settings — YouJiacheng · 2026-08-04