Random-Access Neural Compression of Material Textures

Abstract

The continuous advancement of photorealism in rendering is accompanied by a growth in texture data and, consequently, increasing storage and memory demands. To address this issue, we propose a novel neural compression technique specifically designed for material textures. We unlock two more levels of detail, i.e., 16x more texels, using low bitrate compression, with image quality that is better than advanced image compression techniques, such as AVIF and JPEG XL. At the same time, our method allows on-demand, real-time decompression with random access similar to block texture compression on GPUs, enabling compression on disk and memory. The key idea behind our approach is compressing multiple material textures and their mipmap chains together, and using a small neural network, that is optimized for each material, to decompress them. Finally, we use a custom training implementation to achieve practical compression speeds, whose performance surpasses that of general frameworks, like PyTorch, by an order of magnitude.

22 pages, accepted to ACM SIGGRAPH 2023 Transactions on Graphics

References in corpus (5)

• Instant Neural Graphics Primitives with a Multiresolution Hash Encoding
• Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
• Implicit Neural Representations with Periodic Activation Functions
• Real-time Neural Radiance Caching for Path Tracing
• Passing Multi-Channel Material Textures to a 3-Channel Loss

Cited by in corpus (6)

• Real-Time Neural Appearance Models
• Image-GS: Content-Adaptive Image Representation via 2D Gaussians
• Filtering After Shading With Stochastic Texture Filtering
• Efficient Graphics Representation with Differentiable Indirection
• Neural Product Importance Sampling via Warp Composition
• Improved Stochastic Texture Filtering Through Sample Reuse