Towards 1000-fold Electron Microscopy Image Compression for Connectomics via VQ-VAE with Transformer Prior
Not on the map. Why?
Abstract (arXiv)
Petascale electron microscopy (EM) datasets push storage, transfer, and downstream analysis toward their current limits. We present a vector-quantized variational autoencoder-based (VQ-VAE) compression framework for EM that spans 16x to 1024x and enables pay-as-you-decode usage: top-only decoding for extreme compression, with an optional Transformer prior that predicts bottom tokens (without changing the compression ratio) to restore texture via feature-wise linear modulation (FiLM) and concatenation; we further introduce an ROI-driven workflow that performs selective high-resolution reconstruction from 1024x-compressed latents only where needed.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 3 of its 14 references that are in the map. First posted on arXiv 31 Oct 2025.
Builds on
3 of 14 references.
References · 3
- 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
- 2019Generating Diverse High-Fidelity Images with VQ-VAE-2cited by 1,387
- 2025Budgeted Broadcast: An Activity-Dependent Pruning Rule for Neural Network Efficiencycited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025EMGauss: Continuous Slice-to-3D Reconstruction via Dynamic Gaussian Modeling in Volume Electron Microscopy
- 2026Frequency-Aware Dual-Stream Learning for Balanced Realism and Fidelity in Electron Microscopy Imaging
- 2025Compressing Biology: Evaluating the Stable Diffusion VAE for Phenotypic Drug Discovery
- 2026Structure-Guided Masked Autoencoders for Ultra-High Resolution Scientific Image Understanding
- 2023Reference-Free Isotropic 3D EM Reconstruction using Diffusion Models
- 2026Contrastive Image-Metadata Pre-Training for Materials Transmission Electron Microscopy
- 2023SC-VAE: Sparse Coding-based Variational Autoencoder with Learned ISTA
- 2026SkelEM: Training-Signal Decoupling of Skeleton and Diffusion for Self-supervised Axial Super-Resolution in Volume Microscopy