---
license: mit
task_categories:
- image-classification
language:
- en
tags:
- Multi-Scale
- ImageNet
- Equivariance
size_categories:
- 1K<n<10K
---

# Multi-Scale_ImageNet
The dataset contains multi-scale ImageNet from **Local Scale Equivariance with Latent Deep Equilibrium Canonicalizer** (arXiv:2508.14187) published in ICCV 2025.
The dataset generating pipeline is provided here: https://github.com/ashiq24/local-scale-equivariance/tree/main/datagen_ImNet

## Pipeline Overview

| Global Scaling | Local Scaling |
|----------------|---------------|
| <img src="https://raw.githubusercontent.com/ashiq24/local-scale-equivariance/refs/heads/gh-pages/assets/bird_1_gscale.gif" width="450"/> | <img src="https://raw.githubusercontent.com/ashiq24/local-scale-equivariance/refs/heads/gh-pages/assets/bird_1.gif" width="450"/> |





![image/png](https://cdn-uploads.huggingface.co/production/uploads/661e07e02a8496916011c08a/RPMXOQPeUd4OrqnXvTM5o.png)


If you use this dataset, please consider citing

```
@inproceedings{rahman2025local,
  title={Local Scale Equivariance with Latent Deep Equilibrium Canonicalizer},
  author={Rahman, Md Ashiqur and Yang, Chiao-An and Cheng, Michael N. and Lim, Jun Hao and Jiang, Jeremiah and Lim, Teck-Yian and Yeh, Raymond A.},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2025}
}
```