Finite scalar quantization simplifies VQ-VAE latents by independently rounding a few dimensions to fixed levels, producing an equivalent-sized implicit codebook with competitive performance and no collapse.
Image compression with product quantized masked image modeling
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Factorizing pixel transitions into a learned codebook of patch-level motion primitives, then gating them into latent actions, transfers across morphologies and matches or beats monolithic latent-action baselines in distractor-heavy control tasks.
RDVQ enables joint rate-distortion optimization for vector-quantized generative image compression via differentiable codebook distribution relaxation and an autoregressive entropy model.
citing papers explorer
-
Finite Scalar Quantization: VQ-VAE Made Simple
Finite scalar quantization simplifies VQ-VAE latents by independently rounding a few dimensions to fixed levels, producing an equivalent-sized implicit codebook with competitive performance and no collapse.
-
Latent Actions from Factorized Transition Effects under Agent Ambiguity
Factorizing pixel transitions into a learned codebook of patch-level motion primitives, then gating them into latent actions, transfers across morphologies and matches or beats monolithic latent-action baselines in distractor-heavy control tasks.
-
Differentiable Vector Quantization for Rate-Distortion Optimization of Generative Image Compression
RDVQ enables joint rate-distortion optimization for vector-quantized generative image compression via differentiable codebook distribution relaxation and an autoregressive entropy model.