# A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets

arXiv: https://arxiv.org/abs/1707.08819 · HTML: https://mlsanity.com/paper/1707.08819/

## Key facts

- Cited by 377 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Jul 2017.

## Built on by (most cited first, 30 of 377)

- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference](https://mlsanity.com/paper/2104.01136/) (2021)
- [Multisample Flow Matching: Straightening Flows with Minibatch Couplings](https://mlsanity.com/paper/2304.14772/) (2023)
- [Unlocking High-Accuracy Differentially Private Image Classification through Scale](https://mlsanity.com/paper/2204.13650/) (2022)
- [Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective](https://mlsanity.com/paper/2102.11535/) (2021)
- [Dynamical Regimes of Diffusion Models](https://mlsanity.com/paper/2402.18491/) (2024)
- [Squeeze, Recover and Relabel: Dataset Condensation at ImageNet Scale From A New Perspective](https://mlsanity.com/paper/2306.13092/) (2023)
- [Co-Mixup: Saliency Guided Joint Mixup with Supermodular Diversity](https://mlsanity.com/paper/2102.03065/) (2021)
- [FedALA: Adaptive Local Aggregation for Personalized Federated Learning](https://mlsanity.com/paper/2212.01197/) (2022)
- [POEM: Out-of-Distribution Detection with Posterior Sampling](https://mlsanity.com/paper/2206.13687/) (2022)
- [Dataset Condensation with Contrastive Signals](https://mlsanity.com/paper/2202.02916/) (2022)
- [Elucidating the Exposure Bias in Diffusion Models](https://mlsanity.com/paper/2308.15321/) (2023)
- [Guided Flows for Generative Modeling and Decision Making](https://mlsanity.com/paper/2311.13443/) (2023)
- [How Powerful are Performance Predictors in Neural Architecture Search?](https://mlsanity.com/paper/2104.01177/) (2021)
- [Input Perturbation Reduces Exposure Bias in Diffusion Models](https://mlsanity.com/paper/2301.11706/) (2023)
- [Reflected Diffusion Models](https://mlsanity.com/paper/2304.04740/) (2023)
- [LLMatic: Neural Architecture Search via Large Language Models and Quality Diversity Optimization](https://mlsanity.com/paper/2306.01102/) (2023)
- [Untargeted Backdoor Watermark: Towards Harmless and Stealthy Dataset Copyright Protection](https://mlsanity.com/paper/2210.00875/) (2022)
- [Differentially Private Diffusion Models Generate Useful Synthetic Images](https://mlsanity.com/paper/2302.13861/) (2023)
- [HW-NAS-Bench:Hardware-Aware Neural Architecture Search Benchmark](https://mlsanity.com/paper/2103.10584/) (2021)
- [Probing Representation Forgetting in Supervised and Unsupervised Continual Learning](https://mlsanity.com/paper/2203.13381/) (2022)
- [Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data](https://mlsanity.com/paper/2102.04761/) (2021)
- [AutoMix: Unveiling the Power of Mixup for Stronger Classifiers](https://mlsanity.com/paper/2103.13027/) (2021)
- [Eluding Secure Aggregation in Federated Learning via Model Inconsistency](https://mlsanity.com/paper/2111.07380/) (2021)
- [FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes](https://mlsanity.com/paper/2110.08059/) (2021)
- [Training and Inference on Any-Order Autoregressive Models the Right Way](https://mlsanity.com/paper/2205.13554/) (2022)
- [Your ViT is Secretly a Hybrid Discriminative-Generative Diffusion Model](https://mlsanity.com/paper/2208.07791/) (2022)
- [Flow Matching with General Discrete Paths: A Kinetic-Optimal Perspective](https://mlsanity.com/paper/2412.03487/) (2024)
- [Task-Oriented Communication for Multi-Device Cooperative Edge Inference](https://mlsanity.com/paper/2109.00172/) (2021)
- [Stable, Fast and Accurate: Kernelized Attention with Relative Positional Encoding](https://mlsanity.com/paper/2106.12566/) (2021)

## Abstract (arXiv)

The original ImageNet dataset is a popular large-scale benchmark for training Deep Neural Networks. Since the cost of performing experiments (e.g, algorithm design, architecture search, and hyperparameter tuning) on the original dataset might be prohibitive, we propose to consider a downsampled version of ImageNet. In contrast to the CIFAR datasets and earlier downsampled versions of ImageNet, our proposed ImageNet32×32 (and its variants ImageNet64×64 and ImageNet16×16) contains exactly the same number of classes and images as ImageNet, with the only difference that the images are downsampled to 32×32 pixels per image (64×64 and 16×16 pixels for the variants, respectively). Experiments on these downsampled variants are dramatically faster than on the original ImageNet and the characteristics of the downsampled datasets with respect to optimal hyperparameters appear to remain similar. The proposed datasets and scripts to reproduce our results are available at http://image-net.org/download-images and https://github.com/PatrykChrabaszcz/Imagenet32Scripts