PulseJournal ClubResearchersJournalsExplore
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
September 10, 2025IEEE Transactions on Pattern Analysis and Machine IntelligenceOpen Access

InstructLayout: Instruction-Driven 2D and 3D Layout Synthesis with Semantic Graph Prior

View Full Paper
Ask AI
Bookmark
Share

Authors

CLChenguo LinYLYu-Chen LinPPPanwang Pan

Discussion

Loading...

Member takes

Overview

Generative framework improves control and fidelity in layout synthesis for 2D and 3D tasks, suggesting new benchmarks.

Key Points

  • The framework significantly enhances controllability and fidelity in both 2D and 3D layout synthesis tasks.
  • It outperforms existing state-of-the-art methods by a large margin, particularly in generating layouts from natural language instructions.
  • The integration of a semantic graph prior allows for simultaneous learning of layout appearances and object distributions.
  • New datasets of layout-instruction pairs were curated to benchmark text-driven synthesis against multimodal models.

Cite This Study

Lin et al. (2025) studied this question.

synapsesocial.com/papers/68c1b60d54b1d3bfb60eb225https://doi.org/10.1109/tpami.2025.3595880
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1InstructLayout: Instruction-Driven 2D and 3D Layout Synthesis with Semantic Graph Prior2024
  2. 2Compositional 3D Scene Synthesis with Scene Graph Guided Layout-Shape Generation2024
  3. 3Planner3D: LLM-enhanced Graph Prior Meets 3D Indoor Scene Explicit Regularization2025 · 5 citations
  4. 4OptiScene: LLM-driven Indoor Scene Layout Generation via Scaled Human-aligned Data Synthesis and Multi-Stage Preference Optimization2025
  5. 5CasLayout: Cascaded 3D Layout Diffusion for Indoor Scene Synthesis with Implicit Relation Modeling2026