Research map / Architectures and efficiency / CNN design
Deep Representation-Decoupling Neural Networks for Monaural Music Mixture Separation
Zhuo Li, Hongwei Wang, Miao Zhao, Wenjie Li, Minyi Guo
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 |
|---|---|---|---|
| remove decoupling layer (ablation) | DRDNN · DSD100 vocal | SIR: 6.78 -> 9.39 | Improved |
| remove decoupling layer (ablation) | DRDNN · DSD100 bass | SIR: 2.20 -> 3.14 | Improved |
| remove decoupling layer (ablation) | DRDNN · DSD100 drum | SIR: 5.68 -> 7.35 | Improved |
| remove decoupling layer (ablation) | DRDNN · DSD100 other | SIR: 1.36 -> 1.97 | Improved |
| remove decoupling layer (ablation) | DRDNN · DSD100 | training time: 28% additional time overhead | Improved |
| remove convolution and deconvolution dimension reduction (ablation) | DRDNN · DSD100 vocal | SIR: 9.31 -> 9.39 | No clear effect |
| remove convolution and deconvolution dimension reduction (ablation) | DRDNN · DSD100 drum | SIR: 7.78 -> 7.35 | No clear effect |
| remove decoupling layer (ablation) | DRDNN · training loss MSE | training loss: 0.6 vs 0.003 | Hurt |
| use DRDNN instead of DRNN | DRDNN · DSD100 vocal | SIR: 8.34 -> 9.39 | Improved |
| use DRDNN instead of DRNN | DRDNN · DSD100 drum | SDR: 3.27 -> 3.23 | No clear effect |
| use DRDNN instead of DRNN | DRDNN · DSD100 bass | SIR: 3.26 -> 3.14 | Hurt |
| use DRDNN instead of DRNN | DRDNN · DSD100 other | SAR: 5.64 -> 6.81 | Improved |
Closest papers by meaning
- AutoSchA: Automatic Hierarchical Music Representations via Multi-Relational Node Isolation (AAAI 2026 · similarity 0.85)
- MGE-LDM: Joint Latent Diffusion for Simultaneous Music Generation and Source Extraction (NeurIPS 2025 · similarity 0.83)
- Weakly-supervised Audio Separation via Bi-modal Semantic Similarity (ICLR 2024 · similarity 0.83)
- SoundCount: Sound Counting from Raw Audio with Dyadic Decomposition Neural Network (AAAI 2024 · similarity 0.83)
- Multi-Source Diffusion Models for Simultaneous Music Generation and Separation (ICLR 2024 · similarity 0.83)
- Whole-Song Hierarchical Generation of Symbolic Music Using Cascaded Diffusion Models (ICLR 2024 · similarity 0.83)
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.
