Research map / Architectures and efficiency / CNN design
TI-Pooling: Transformation-Invariant Pooling for Feature Learning in Convolutional Neural Networks
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 |
|---|---|---|---|
| Use TI-pooling over 24 rotations instead of a TIRBM baseline | MNIST rotation classification · mnist-rot-12k | error (%): 4.2 -> 1.2 | Helped |
| Use TI-pooling instead of STN (affine) | Half-rotated MNIST · Half-rotated MNIST | error (%): 0.7 -> 0.8 | Hurt |
| Use TI-pooling instead of CNN with data augmentation | Neuronal segmentation · Neuronal segmentation test | pixel error (%): 8.1 -> 7.0 | Helped |
| Use TI-pooling instead of MIL over CNN | Neuronal segmentation · Neuronal segmentation test | pixel error (%): 8.9 -> 7.0 | Helped |
| Add dropout (ablation) | TI-pooling · Neuronal segmentation test | pixel error (%): 7.4 -> 7.0 | Helped |
| Increase the number of sampled transformations in the TI-pooling set (ablation) | TI-pooling · Neuronal segmentation validation | validation error | Helped |
Closest papers by meaning
- Detail-Preserving Pooling in Deep Networks (CVPR 2018 · similarity 0.88)
- S3Pool: Pooling With Stochastic Spatial Sampling (CVPR 2017 · similarity 0.87)
- Gaussian-Based Pooling for Convolutional Neural Networks (NeurIPS 2019 · similarity 0.87)
- Inductive Bias of Deep Convolutional Networks through Pooling Geometry (ICLR 2017 · similarity 0.87)
- Mind the Pool: Convolutional Neural Networks Can Overfit Input Size (ICLR 2023 · similarity 0.87)
- Beyond Skip Connection: Pooling and Unpooling Design for Elimination Singularities (AAAI 2025 · 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.
