CryoFastAR: Fast Cryo-EM Ab Initio Reconstruction Made Easy
By: Jiakai Zhang , Shouchen Zhou , Haizhao Dai and more
Potential Business Impact:
Helps scientists see tiny protein shapes faster.
Pose estimation from unordered images is fundamental for 3D reconstruction, robotics, and scientific imaging. Recent geometric foundation models, such as DUSt3R, enable end-to-end dense 3D reconstruction but remain underexplored in scientific imaging fields like cryo-electron microscopy (cryo-EM) for near-atomic protein reconstruction. In cryo-EM, pose estimation and 3D reconstruction from unordered particle images still depend on time-consuming iterative optimization, primarily due to challenges such as low signal-to-noise ratios (SNR) and distortions from the contrast transfer function (CTF). We introduce CryoFastAR, the first geometric foundation model that can directly predict poses from Cryo-EM noisy images for Fast ab initio Reconstruction. By integrating multi-view features and training on large-scale simulated cryo-EM data with realistic noise and CTF modulations, CryoFastAR enhances pose estimation accuracy and generalization. To enhance training stability, we propose a progressive training strategy that first allows the model to extract essential features under simpler conditions before gradually increasing difficulty to improve robustness. Experiments show that CryoFastAR achieves comparable quality while significantly accelerating inference over traditional iterative approaches on both synthetic and real datasets.
Similar Papers
GEM: 3D Gaussian Splatting for Efficient and Accurate Cryo-EM Reconstruction
CV and Pattern Recognition
Makes seeing tiny things faster and clearer.
Multiscale guidance of AlphaFold3 with heterogeneous cryo-EM data
Machine Learning (CS)
Helps scientists see how moving body parts work.
Cryo-EM as a Stochastic Inverse Problem
Machine Learning (Stat)
Shows how tiny parts of bodies move.