Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Learned Compression for Compressed Learning

About

Modern sensors produce increasingly rich streams of high-resolution data. Due to resource constraints, machine learning systems discard the vast majority of this information via resolution reduction. Compressed-domain learning allows models to operate on compact latent representations, allowing higher effective resolution for the same budget. However, existing compression systems are not ideal for compressed learning. Linear transform coding and end-to-end learned compression systems reduce bitrate, but do not uniformly reduce dimensionality; thus, they do not meaningfully increase efficiency. Generative autoencoders reduce dimensionality, but their adversarial or perceptual objectives lead to significant information loss. To address these limitations, we introduce WaLLoC (Wavelet Learned Lossy Compression), a neural codec architecture that combines linear transform coding with nonlinear dimensionality-reducing autoencoders. WaLLoC sandwiches a shallow, asymmetric autoencoder and entropy bottleneck between an invertible wavelet packet transform. Across several key metrics, WaLLoC outperforms the autoencoders used in state-of-the-art latent diffusion models. WaLLoC does not require perceptual or adversarial losses to represent high-frequency detail, providing compatibility with modalities beyond RGB images and stereo audio. WaLLoC's encoder consists almost entirely of linear operations, making it exceptionally efficient and suitable for mobile computing, remote sensing, and learning directly from compressed data. We demonstrate WaLLoC's capability for compressed-domain learning across several tasks, including image classification, colorization, document understanding, and music source separation. Our code, experiments, and pre-trained audio and image codecs are available at https://ut-sysml.org/walloc

Dan Jacobellis, Neeraja J. Yadwadkar• 2024

Related benchmarks

TaskDatasetResultRank
Image CompressionKodak
Bits Per Pixel (bpp)0.0097
31
Image CompressionKodak 512 × 768 and 768 × 512
Bits Per Pixel (bpp)0.6171
16
Image CompressionImageNet-1k 224 × 224
bpp0.772
16
Image CompressionKodak
BD-Rate (PSNR)-27.61
4
Image CompressionKodak < 0.0464 bpp regime
bpp0.0204
3
Showing 5 of 5 rows

Other info

Follow for update