AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

Scaling Backwards: Minimal Synthetic Pre-training?

Ryo Nakamura, Ryu Tadokoro, Ryosuke Yamada, Yuki M Asano, Iro Laina, Christian Rupprecht, Nakamasa Inoue, Rio Yokota, Hirokatsu Kataoka

ECCV 2024 · Architectures and efficiency › Pruning and efficient tuning · approach: Efficient ViT adaptation · 5 citations (OpenAlex)

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 changedWhereMeasuredEffect
pre-train on a single fractal image with LPCE loss instead of 1M-image FractalDBViT-Tiny · CIFAR-100 fine-tuningtop-1 accuracy: 81.6 -> 84.2Improved
pre-train on ImageNet-21k scale versus single fractal 1p-frac for ViT-BViT-B · ImageNet-1k fine-tuningtop-1 accuracy: 81.8 -> 82.1Improved
use single Gaussian or uniform noise image with LPCE instead of a fractal (ablation)ViT-Tiny · CIFAR-100 fine-tuningtop-1 accuracy: 84.2 -> 1.1Hurt
reduce perturbation degree Delta from 0.05 to 0.01 (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 83.0 -> 19.9Hurt
increase perturbation degree beyond 0.1 (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 84.2 -> 82.6Hurt
use the most complex fractal shape (sigma 3.5) instead of simpler ones (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 81.3 -> 84.2Improved
increase numerical integration sampling points L from 16 to 1000 (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 78.7 -> 84.2Improved
remove random crop from pre-training augmentation (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 84.2 -> 80.1Hurt
remove Mixup and Cutmix from pre-training augmentation (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 84.2 -> 80.4Hurt
remove horizontal flipping from pre-training augmentation (ablation)1p-frac · CIFAR-100 fine-tuningtop-1 accuracy: 84.2 -> 84.3No clear effect
use affine perturbation instead of elastic perturbation for single real image (ablation)single real image pre-training · CIFAR-100 (mean of five images)mean accuracy: 37.0 -> 81.9Improved
use Canny edge image instead of RGB image for single real image pre-training (ablation)single real image pre-training · CIFAR-100 (mean of five images)mean accuracy: 81.6 -> 82.2Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy