Research map / 3D vision / Depth and stereo
Robust Signal Enhancement via Fractional Detail Views and Knowledge Guided Multi-view Fusion
Zikun Jin, Yuhua Qian, Xinyan Liang, Jiaqian Zhang, Haijun Geng
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 |
|---|---|---|---|
| replace standard convolution with LRF view (ablation) | FracKGMF · VoiceBank+DEMAND | PESQ: 3.14 -> 3.21 | Improved |
| replace standard convolution with FracConv FD view (ablation) | FracKGMF · VoiceBank+DEMAND | PESQ: 3.14 -> 3.25 | Improved |
| fuse LRF and FD views by naive summation (ablation) | FracKGMF · VoiceBank+DEMAND | PESQ: 3.25 (FD only) -> 3.25 (sum) | No clear effect |
| use Wiener-guided KGMF fusion instead of summation (ablation) | FracKGMF · VoiceBank+DEMAND | PESQ: 3.25 -> 3.32 | Improved |
| use Wiener-guided KGMF fusion instead of summation (ablation) | FracKGMF · VoiceBank+DEMAND | COVL: 3.97 -> 4.04 | Improved |
| vary FracConv stabilization constant epsilon over three orders of magnitude (ablation) | FracConv · VoiceBank+DEMAND | PESQ: 3.30 / 3.29 / 3.32 | No clear effect |
| FracKGMF vs strongest baselines at extremely low SNR | EM enhancement · Rydberg EM at -20 dB | mean SNR improvement (dB): 34.77 dB mean improvement (4-bins); 33.04 dB across three EM datasets | Improved |
| FracKGMF vs MFSE on SSNR | speech enhancement · VoiceBank+DEMAND | SSNR: 10.63 (MFSE) vs 10.37 | Hurt |
Closest papers by meaning
- Signal Enhancement via Multi-view Dynamic Representation and Alignment-aware Fusion (AAAI 2026 · similarity 0.91)
- Dual-View Predictive Diffusion: Lightweight Speech Enhancement via Spectrogram-Image Synergy (ICML 2026 · similarity 0.83)
- Fast Fourier Convolution (NeurIPS 2020 · similarity 0.83)
- Learned Image Compression via Sparse Attention and Adaptive Frequency (CVPR 2026 · similarity 0.83)
- Query-Based Asymmetric Modeling with Decoupled Input–Output Rates for Speech Restoration (ICML 2026 · similarity 0.83)
- BA-SAM: Scalable Bias-Mode Attention Mask for Segment Anything Model (CVPR 2024 · similarity 0.81)
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.
