Multi-Part Object Representations via Graph Structures and Co-Part Discovery

2025/12/20 by Alex Foo, Wynne Hsu, Foo, Alex +3
Computer Science · #Advanced Neural Network Applications #Domain Adaptation and Few-Shot Learning #Graph Theory and Algorithms

paper · doi:10.48550/arxiv.2512.18192

Abstract

Discovering object-centric representations from images can significantly enhance the robustness, sample efficiency and generalizability of vision models. Works on images with multi-part objects typically follow an implicit object representation approach, which fail to recognize these learned objects in occluded or out-of-distribution contexts. This is due to the assumption that object part-whole relations are implicitly encoded into the representations through indirect training objectives. We address this limitation by proposing a novel method that leverages on explicit graph representations for parts and present a co-part object discovery algorithm. We then introduce three benchmarks to evaluate the robustness of object-centric methods in recognizing multi-part objects within occluded and out-of-distribution settings. Experimental results on simulated, realistic, and real-world images show marked improvements in the quality of discovered objects compared to state-of-the-art methods, as well as the accurate recognition of multi-part objects in occluded and out-of-distribution contexts. We also show that the discovered object-centric representations can more accurately predict key object properties in a downstream task, highlighting the potential of our method to advance the field of object-centric representations.

Citations

• SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models
• Micrograph segmentations for DDEVD
• Object-Centric Slot Diffusion
• Cut and Learn for Unsupervised Object Detection and Instance Segmentation
• Neural Systematic Binder
• Improving Object-centric Learning with Query Optimization
• Promising or Elusive? Unsupervised Object Segmentation from Real-world Single Images
• Bridging the Gap to Real-World Object-Centric Learning
• Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household Items
• Compositional Scene Representation Learning via Reconstruction: A Survey
• Conditional Object-Centric Learning from Video
• ClevrTex: A Texture-Rich Benchmark for Unsupervised Multi-Object Segmentation
• Unsupervised Part Discovery from Contrastive Reconstruction
• Illiterate DALL-E Learns to Compose
• Generalization and Robustness Implications in Object-Centric Learning
• Efficient Iterative Amortized Inference for Learning Symmetric and Disentangled Multi-Object Representations
• Toward Causal Representation Learning
• Emerging Properties in Self-Supervised Vision Transformers
• GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement
• On the Binding Problem in Artificial Neural Networks
• Object-Centric Learning with Slot Attention
• End-to-End Object Detection with Transformers
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representations
• MONet: Unsupervised Scene Decomposition and Representation
• Deep Object Co-Segmentation
• Mask R-CNN
• Microsoft COCO: Common Objects in Context
• SLIC Superpixels Compared to State-of-the-Art Superpixel Methods

Related

• Unsupervised Part Discovery via Dual Representation Alignment
• Unsupervised Part Discovery from Contrastive Reconstruction
• Toward Parts-Based Scene Understanding with Pixel-Support Parts-Sparse Pictorial Structures
• Object-centric Learning with Cyclic Walks between Parts and Whole
• Learning Object-Centric Representations of Multi-Object Scenes from Multiple Views