Research map / Optimization / Nonconvex and smooth optimization
Reshaped Wirtinger Flow for Solving Quadratic System of Equations
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use magnitude loss (|a^T z|-y)^2 instead of squared-intensity WF loss | reshaped-WF · Real Gaussian n=1000 m=8n | iterations: 319.2 -> 72 | Helped |
| use magnitude loss instead of WF loss | reshaped-WF · Real Gaussian n=1000 m=8n | time (s): 2.104 -> 0.477 | Helped |
| use reshaped-WF instead of truncated-WF | reshaped-WF · Complex Gaussian n=1000 m=8n | iterations: 486.7 -> 272.7 | Helped |
| use reshaped-WF instead of truncated-WF | reshaped-WF · Galaxy image CDP L=16 | time (s): 567 -> 141 | Helped |
| use reshaped-WF instead of AltMinPhase | reshaped-WF · Real Gaussian n=1000 m=8n | iterations: 5.8 -> 72 | Hurt |
| use reshaped-WF instead of AltMinPhase | reshaped-WF · Real Gaussian n=1000 m=8n | time (s): 0.908 -> 0.477 | Helped |
| use reshaped-WF instead of truncated-WF | reshaped-WF · Real Gaussian sample complexity | success rate | No clear effect |
| use reshaped-WF instead of truncated-WF/WF | reshaped-WF · Complex and CDP sample complexity | success rate | Helped |
| use magnitude-based spectral initialization with two-sided truncation | reshaped-WF · Gaussian m=6n | initial relative error | Helped |
| use reshaped-WF instead of truncated-WF under Poisson noise | reshaped-WF · Poisson noise | relative error | Helped |
Closest papers by meaning
- Provable Non-convex Phase Retrieval with Outliers: Median TruncatedWirtinger Flow (ICML 2016 · similarity 0.87)
- Hessian-guided Perturbed Wasserstein Gradient Flows for Escaping Saddle Points (NeurIPS 2025 · similarity 0.83)
- Semi-Random Matrix Completion via Flow-Based Adaptive Reweighting (NeurIPS 2024 · similarity 0.83)
- On Fitting Flow Models with Large Sinkhorn Couplings (NeurIPS 2026 · similarity 0.82)
- Flow Matching from Viewpoint of Proximal Operators (NeurIPS 2026 · similarity 0.82)
- CDFlow: Building Invertible Layers with Circulant and Diagonal Matrices (NeurIPS 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.