Copyright-Aware Incentive Scheme for Generative Art Models Using Hierarchical Reinforcement Learning

TL;DR

This is the first technical work that introduces to incentive contributors and protect their copyrights by compensating them and a hierarchical budget allocation method based on reinforcement learning to determine the budget for each round and the remuneration of the data holder based on the data holder's contribution and copyright loss in each round.

Abstract

Generative art using Diffusion models has achieved remarkable performance in image generation and text-to-image tasks. However, the increasing demand for training data in generative art raises significant concerns about copyright infringement, as models can produce images highly similar to copyrighted works. Existing solutions attempt to mitigate this by perturbing Diffusion models to reduce the likelihood of generating such images, but this often compromises model performance. Another approach focuses on economically compensating data holders for their contributions, yet it fails to address copyright loss adequately. Our approach begin with the introduction of a novel copyright metric grounded in copyright law and court precedents on infringement. We then employ the TRAK method to estimate the contribution of data holders. To accommodate the continuous data collection process, we divide the training into multiple rounds. Finally, We designed a hierarchical budget allocation method based on reinforcement learning to determine the budget for each round and the remuneration of the data holder based on the data holder's contribution and copyright loss in each round. Extensive experiments across three datasets show that our method outperforms all eight benchmarks, demonstrating its effectiveness in optimizing budget distribution in a copyright-aware manner. To the best of our knowledge, this is the first technical work that introduces to incentive contributors and protect their copyrights by compensating them.

Figures (5)

• Figure 1: The Results of the Spearman's Rank Correlation Test for Copyright Loss and Contribution indicate distinct group behaviors: Points in group A exhibit small copyright loss and significant contributions, while points in group B show minimal contribution but substantial copyright loss.
• Figure 2: Workflow of Computation of Copyright Loss.
• Figure 3: Workflow of Hierarchical Reinforcement Learning. The outer RL distributes the budget $B^1, B^2, ..., B^T$ to each round. The inner RL distributes the budget to each data holder according to the contribution and the copyright loss, and the model with be trained with the corresponding data. The quality of the output model serves as the reward of the outer RL.
• Figure 4: Comparison of Our Method with the Baselines on Three Datasets.
• Figure 5: Ablation Results on Three Datasets.

Theorems & Definitions (2)

• definition 1: Text-to-image diffusion models
• definition 2: Data attribution method: TRAK