MINR: Efficient Implicit Neural Representations for Multi-Image Encoding
This addresses a bottleneck in multi-image encoding for INR applications, offering a more efficient solution for tasks like image reconstruction and super-resolution, though it is incremental as it builds on existing INR methods.
The paper tackles the computational and storage inefficiency of using separate neural networks for each image in Implicit Neural Representations (INRs) by proposing MINR, which shares intermediate layers across multiple images, saving up to 60% parameters while maintaining comparable performance, such as an average PSNR of 34 dB for 100 images.
Implicit Neural Representations (INRs) aim to parameterize discrete signals through implicit continuous functions. However, formulating each image with a separate neural network~(typically, a Multi-Layer Perceptron (MLP)) leads to computational and storage inefficiencies when encoding multi-images. To address this issue, we propose MINR, sharing specific layers to encode multi-image efficiently. We first compare the layer-wise weight distributions for several trained INRs and find that corresponding intermediate layers follow highly similar distribution patterns. Motivated by this, we share these intermediate layers across multiple images while preserving the input and output layers as input-specific. In addition, we design an extra novel projection layer for each image to capture its unique features. Experimental results on image reconstruction and super-resolution tasks demonstrate that MINR can save up to 60\% parameters while maintaining comparable performance. Particularly, MINR scales effectively to handle 100 images, maintaining an average peak signal-to-noise ratio (PSNR) of 34 dB. Further analysis of various backbones proves the robustness of the proposed MINR.