OID-PPO: Optimal Interior Design using Proximal Policy Optimization by Transforming Design Guidelines into Reward Functions

2025-08-01

AI Summary

Residential interior layout design suffers from unstructured configurations, high computational cost, heavy reliance on expert knowledge, and existing approaches—optimization-based, deep learning–based, or reinforcement learning–based—struggle to jointly satisfy functional requirements and aesthetic quality. Method: This paper proposes a rule-guided reinforcement learning framework operating in continuous action space. It explicitly encodes multi-dimensional design principles into a structured reward function and employs Proximal Policy Optimization (PPO) with a diagonal Gaussian policy network to enable flexible, physically plausible furniture placement while addressing partial observability. Contribution/Results: Experiments demonstrate significant improvements over baselines across diverse room geometries and furniture sets, yielding higher-quality layouts and enhanced computational efficiency. Ablation studies confirm the effectiveness of each incorporated design constraint.

Technology Category

• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Designing residential interiors strongly impacts occupant satisfaction but remains challenging due to unstructured spatial layouts, high computational demands, and reliance on expert knowledge. Existing methods based on optimization or deep learning are either computationally expensive or constrained by data scarcity. Reinforcement learning (RL) approaches often limit furniture placement to discrete positions and fail to incorporate design principles adequately. We propose OID-PPO, a novel RL framework for Optimal Interior Design using Proximal Policy Optimization, which integrates expert-defined functional and visual guidelines into a structured reward function. OID-PPO utilizes a diagonal Gaussian policy for continuous and flexible furniture placement, effectively exploring latent environmental dynamics under partial observability. Experiments conducted across diverse room shapes and furniture configurations demonstrate that OID-PPO significantly outperforms state-of-the-art methods in terms of layout quality and computational efficiency. Ablation studies further demonstrate the impact of structured guideline integration and reveal the distinct contributions of individual design constraints.

Problem

• Optimizing residential interior design for occupant satisfaction
• Overcoming computational and data constraints in design methods
• Enhancing furniture placement with continuous RL and design principles

Innovation

• Uses Proximal Policy Optimization for interior design
• Integrates design guidelines into reward functions
• Employs diagonal Gaussian policy for flexible placement