Affogato: Learning Open-Vocabulary Affordance Grounding with Automated Data Generation at Scale

TL;DR

This paper introduces Affogato, a large-scale dataset and models for open-vocabulary affordance grounding, enabling better understanding of object interactions through natural language descriptions and heatmap localization.

Contribution

The paper presents a new large-scale benchmark dataset with 150K instances for affordance grounding and develops vision-language models that leverage this dataset for improved open-vocabulary and cross-domain performance.

Findings

Models trained on Affogato outperform existing benchmarks.

The dataset enables effective open-vocabulary affordance grounding.

Models demonstrate strong cross-domain generalization.

Abstract

Affordance grounding-localizing object regions based on natural language descriptions of interactions-is a critical challenge for enabling intelligent agents to understand and interact with their environments. However, this task remains challenging due to the need for fine-grained part-level localization, the ambiguity arising from multiple valid interaction regions, and the scarcity of large-scale datasets. In this work, we introduce Affogato, a large-scale benchmark comprising 150K instances, annotated with open-vocabulary text descriptions and corresponding 3D affordance heatmaps across a diverse set of objects and interactions. Building on this benchmark, we develop simple yet effective vision-language models that leverage pretrained part-aware vision backbones and a text-conditional heatmap decoder. Our models trained with the Affogato dataset achieve promising performance on the existing 2D and 3D benchmarks, and notably, exhibit effectiveness in open-vocabulary cross-domain generalization. The Affogato dataset is shared in public: https://huggingface.co/datasets/project-affogato/affogato