Research map / Label-efficient and robust learning / Recognition and long-tail
Deep Spectral Clustering Using Dual Autoencoder Network
Check 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 |
|---|---|---|---|
| Replace ConvAE with mutual information estimation (ablation) | Convolutional autoencoder · MNIST-test | NMI: 0.751 -> 0.796 | Helped |
| Replace ConvAE with mutual information estimation (ablation) | Convolutional autoencoder · MNIST-test | ACC: 0.781 -> 0.844 | Helped |
| Use improved reconstruction loss instead of original reconstruction loss (ablation) | Convolutional autoencoder · MNIST-test | NMI: 0.751 -> 0.801 | Helped |
| Use improved reconstruction loss instead of original reconstruction loss (ablation) | Convolutional autoencoder · MNIST-test | ACC: 0.781 -> 0.850 | Helped |
| Add mutual information estimation and improved reconstruction loss (ablation) | Convolutional autoencoder · MNIST-test | NMI: 0.751 -> 0.914 | Helped |
| Add mutual information estimation and improved reconstruction loss (ablation) | Convolutional autoencoder · MNIST-test | ACC: 0.781 -> 0.961 | Helped |
| Add joint deep spectral clustering (ablation) | ConvAE+MI+RS · MNIST-test | NMI: 0.914 -> 0.946 | Helped |
| Add joint deep spectral clustering (ablation) | ConvAE+MI+RS · MNIST-test | ACC: 0.961 -> 0.980 | Helped |
| Use the full method rather than the strongest listed baseline | Clustering methods · USPS | ACC: 0.950 -> 0.869 | Hurt |
| Use the full method rather than the strongest listed baseline | Clustering methods · USPS | NMI: 0.913 -> 0.857 | Hurt |
| Use the full method rather than the strongest listed baseline | Clustering methods · MNIST-full | ACC: 0.965 -> 0.978 | Helped |
| Use the full method rather than the strongest listed baseline | Clustering methods · MNIST-test | ACC: 0.963 -> 0.980 | Helped |
Closest papers by meaning
- SpectralNet: Spectral Clustering using Deep Neural Networks (ICLR 2018 · similarity 0.93)
- Deep Spectral Clustering Learning (ICML 2017 · similarity 0.90)
- Clustering-friendly Representation Learning via Instance Discrimination and Feature Decorrelation (ICLR 2021 · similarity 0.89)
- Deep Subspace Clustering Networks (NeurIPS 2017 · similarity 0.88)
- Deep Clustering Based on Sparse Kolmogorov-Arnold Network and Spectral Constraint (AAAI 2026 · similarity 0.88)
- Exploring a Principled Framework for Deep Subspace Clustering (ICLR 2025 · similarity 0.88)
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.