# IBCapsNet: Information Bottleneck Capsule Network for Noise-Robust Representation Learning

arXiv: https://arxiv.org/abs/2603.20682 · HTML: https://mlsanity.com/paper/2603.20682/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 19 references that are in the map.
- First posted on arXiv 21 Mar 2026.

## Builds on (most cited first, 8 of 8)

- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [Deep Variational Information Bottleneck](https://mlsanity.com/paper/1612.00410/) (2016)
- [Dynamic Routing Between Capsules](https://mlsanity.com/paper/1710.09829/) (2017)
- [Graph Information Bottleneck](https://mlsanity.com/paper/2010.12811/) (2020)
- [Attention-Based Capsule Networks with Dynamic Routing for Relation Extraction](https://mlsanity.com/paper/1812.11321/) (2018)
- [Information Bottleneck and its Applications in Deep Learning](https://mlsanity.com/paper/1904.03743/) (2019)
- [OrthCaps: An Orthogonal CapsNet with Sparse Attention Routing and Pruning](https://mlsanity.com/paper/2403.13351/) (2024)
- [Is the Information Bottleneck Robust Enough? Towards Label-Noise Resistant Information Bottleneck Learning](https://mlsanity.com/paper/2512.10573/) (2025)

## Abstract (arXiv)

Capsule networks (CapsNets) are superior at modeling hierarchical spatial relationships but suffer from two critical limitations: high computational cost due to iterative dynamic routing and poor robustness under input corruptions. To address these issues, we propose IBCapsNet, a novel capsule architecture grounded in the Information Bottleneck (IB) principle. Instead of iterative routing, IBCapsNet employs a one-pass variational aggregation mechanism, where primary capsules are first compressed into a global context representation and then processed by class-specific variational autoencoders (VAEs) to infer latent capsules regularized by the KL divergence. This design enables efficient inference while inherently filtering out noise. Experiments on MNIST, Fashion-MNIST, SVHN and CIFAR-10 show that IBCapsNet matches CapsNet in clean-data accuracy (achieving 99.41% on MNIST and 92.01% on SVHN), yet significantly outperforms it under four types of synthetic noise - demonstrating average improvements of +17.10% and +14.54% for clamped additive and multiplicative noise, respectively. Moreover, IBCapsNet achieves 2.54x faster training and 3.64x higher inference throughput compared to CapsNet, while reducing model parameters by 4.66%. Our work bridges information-theoretic representation learning with capsule networks, offering a principled path toward robust, efficient, and interpretable deep models. Code is available at https://github.com/cxiang26/IBCapsnet