Research map / 3D vision / Depth and stereo
Physics-Driven Neural Compensation for Electrical Impedance Tomography
Chuyu Wang, Huiting Deng, Dong Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add sensitivity-based neural compensation (level mapping) to hybrid representation (ablation) | PhyNC · Case 2 ablation (qualitative) | reconstruction quality | Improved |
| add Fourier features to Mip-Map embedding (ablation) | PhyNC · Case 2 ablation (qualitative) | reconstruction quality | Improved |
| add global feature (ablation) | PhyNC · Case 2 ablation (qualitative) | background consistency | Improved |
| add frequency regularization with stochastic resampling (ablation) | PhyNC · Case 2 ablation (qualitative) | reconstruction quality | Improved |
| zero out Fourier features or global feature at test (ablation) | PhyNC · Cases 2 and 3 | reconstructed shape | No clear effect |
| zero out Mip-Map embedding (ablation) | PhyNC · Cases 2 and 3 | reconstructed shape | Hurt |
| use PhyNC instead of Hash embedding | Hash · Simulated Case 3 | PSNR: 24.62 -> 29.28 | Improved |
| add sensitivity compensation on top of Hybhash | Hybhash · Simulated Case 2 | PSNR: 23.89 -> 27.73 | Improved |
| use neural field (FFP) instead of TV regularization | TV · Simulated Case 1 | PSNR: 22.39 -> 24.10 | Improved |
| increase FE mesh to 3,154 nodes (ablation) | PhyNC · FE mesh density sweep | PSNR, SSIM, LPIPS | Improved |
| increase FE mesh to finest 5,833 nodes (ablation) | all neural field methods · FE mesh density sweep | PSNR, SSIM, LPIPS | Hurt |
| add sensitivity compensation to improve robustness to mesh density | PhyNC · FE mesh density sweep | limiting node count: 1,630 -> 3,928 | Improved |
Closest papers by meaning
- DeepEIT: Deep Image Prior Enabled Electrical Impedance Tomography (TPAMI 2023 · similarity 0.90)
- Improved Training of Physics-Informed Neural Networks Using Energy-Based Priors: a Study on Electrical Impedance Tomography (ICLR 2023 · similarity 0.89)
- Transformer Meets Boundary Value Inverse Problems (ICLR 2023 · similarity 0.86)
- Neural Field Thermal Tomography: A Differentiable Physics Framework for Non-Destructive Evaluation (NeurIPS 2026 · similarity 0.82)
- Towards Prospective Medical Image Reconstruction via Knowledge-Informed Dynamic Optimal Transport (NeurIPS 2025 · similarity 0.80)
- Boosting ViT-based MRI Reconstruction from the Perspectives of Frequency Modulation, Spatial Purification, and Scale Diversification (AAAI 2025 · similarity 0.80)
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.
