Research map / Probabilistic and scientific modeling / Variational and generative sampling
Understanding and Improving Interpolation in Autoencoders via an Adversarial Regularizer
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 |
|---|---|---|---|
| Add adversarial interpolation regularization | Baseline autoencoder · Synthetic lines | Mean Distance (× 10⁻³): 6.88$\pm$0.21 -> 0.24$\pm$0.01 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · Synthetic lines | Smoothness: 0.44$\pm$0.04 -> 0.10$\pm$0.01 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · MNIST, d_z=32 | single-layer classifier accuracy: 94.90$\pm$0.14 -> 98.25$\pm$0.11 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · MNIST, d_z=256 | single-layer classifier accuracy: 93.94$\pm$0.13 -> 99.00$\pm$0.08 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · SVHN, d_z=32 | single-layer classifier accuracy: 26.21$\pm$0.42 -> 34.47$\pm$1.14 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · SVHN, d_z=256 | single-layer classifier accuracy: 22.74$\pm$0.05 -> 85.14$\pm$0.20 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · CIFAR-10, d_z=256 | single-layer classifier accuracy: 47.92$\pm$0.20 -> 52.77$\pm$0.45; Denoising: 53.78$\pm$0.36 | Mixed |
| Add adversarial interpolation regularization | Baseline autoencoder · CIFAR-10, d_z=1024 | single-layer classifier accuracy: 51.62$\pm$0.25 -> 63.99$\pm$0.47 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · MNIST, d_z=32 | K-Means clustering accuracy: 77.56 -> 94.38 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · MNIST, d_z=256 | K-Means clustering accuracy: 53.70 -> 96.17 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · SVHN, d_z=32 | K-Means clustering accuracy: 19.38 -> 20.86 | Helped |
| Add adversarial interpolation regularization | Baseline autoencoder · SVHN, d_z=256 | K-Means clustering accuracy: 15.62 -> 24.98 | Helped |
Closest papers by meaning
- Autoencoder Image Interpolation by Shaping the Latent Space (ICML 2021 · similarity 0.91)
- Supervised autoencoders: Improving generalization performance with unsupervised regularizers (NeurIPS 2018 · similarity 0.88)
- Regularized linear autoencoders recover the principal components, eventually (NeurIPS 2020 · similarity 0.86)
- Vector Quantization-Based Regularization for Autoencoders (AAAI 2020 · similarity 0.86)
- Adversarially Robust Representations with Smooth Encoders (ICLR 2020 · similarity 0.86)
- Learning Autoencoders with Relational Regularization (ICML 2020 · similarity 0.86)
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.