Research map / Probabilistic and scientific modeling / Variational and generative sampling
Mapping Networks
Lord Sen, Shyamapada Mukherjee
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 the trainable target network with a mapping network (latent vector) on CNN1 | CNN1 · FMNIST | accuracy: 92.89% -> 93.91% | Improved |
| replace trainable CNN2 with mapping network (single latent vector, 2048) | CNN2 · MNIST | accuracy: 98.69% -> 98.66% | No clear effect |
| use layer-wise latent vectors instead of a single latent vector | Mapping Network on CNN2 · FMNIST | accuracy: 91.88% -> 93.35% | Improved |
| mapping network replaces CNN2 for deepfake detection | CNN2 · Celeb-DF | accuracy: 79.03% -> 85.90% | Improved |
| mapping network replaces CNN3 on segmentation | CNN3 · Cityscapes | mIoU: 0.4957 -> 0.4623 (pixel acc 93.21% -> 97.92%) | Mixed |
| mapping network fine-tunes ResNet50 on all layers | ResNet50 · Celeb-DF | accuracy: 95.23% -> 95.10% | No clear effect |
| add weight modulation to the mapping weights (ablation) | Ours* · FMNIST | accuracy: 87.66% -> 91.88% | Improved |
| make mapping weights fully trainable (ablation) | Ours* · FMNIST | accuracy: 91.88% -> 90.11% | Hurt |
| add stability, smoothness and alignment terms to the mapping loss (ablation) | Ours* (2048) · FMNIST | accuracy: 87.88% -> 91.88% | Improved |
| add low-rank decomposition (ablation) | Ours* · FMNIST | accuracy: 91.88% -> 90.67% | Hurt |
| add 90% unstructured pruning (ablation) | Ours* · FMNIST | accuracy: 91.88% -> 88.70% | Hurt |
Closest papers by meaning
- Deep Pyramidal Residual Networks (CVPR 2017 · similarity 0.86)
- On Local Overfitting and Forgetting in Deep Neural Networks (AAAI 2025 · similarity 0.86)
- Modern Neural Networks Generalize on Small Data Sets (NeurIPS 2018 · similarity 0.86)
- Proximal Mapping for Deep Regularization (NeurIPS 2020 · similarity 0.86)
- More is Better: when Infinite Overparameterization is Optimal and Overfitting is Obligatory (ICLR 2024 · similarity 0.85)
- Efficient Representation of Low-Dimensional Manifolds using Deep Networks (ICLR 2017 · similarity 0.85)
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.
