Generalizable Implicit Neural Representations via Instance Pattern Composers
This addresses the problem of generalizable INRs for researchers and practitioners in machine learning, offering a novel method that is compatible with existing techniques like meta-learning and hypernetworks.
The paper tackles the challenge of enabling implicit neural representations (INRs) to generalize across data instances by introducing a framework that modulates a small set of weights in an early MLP layer as an instance pattern composer, achieving high performance on domains like audio, image, and 3D objects.
Despite recent advances in implicit neural representations (INRs), it remains challenging for a coordinate-based multi-layer perceptron (MLP) of INRs to learn a common representation across data instances and generalize it for unseen instances. In this work, we introduce a simple yet effective framework for generalizable INRs that enables a coordinate-based MLP to represent complex data instances by modulating only a small set of weights in an early MLP layer as an instance pattern composer; the remaining MLP weights learn pattern composition rules for common representations across instances. Our generalizable INR framework is fully compatible with existing meta-learning and hypernetworks in learning to predict the modulated weight for unseen instances. Extensive experiments demonstrate that our method achieves high performance on a wide range of domains such as an audio, image, and 3D object, while the ablation study validates our weight modulation.