SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.15316 · 2023

Training on Thin Air: Improve Image Classification with Generated Data

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 9 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
yongchao97/diffusion_inversion canonical 4 of 9
FunctionStatusWhere it lives
EfficientNetB0 Ran yongchao97/diffusion_inversion/src/models/efficientnet.py
code served (permissive licence) · get_code("f7b96dbdddb56a68")
drop_connect Ran yongchao97/diffusion_inversion/src/models/efficientnet.py
code served (permissive licence) · get_code("4304a326c593f8db")
get_lr Ran yongchao97/diffusion_inversion/src/train_net.py
code served (permissive licence) · get_code("7337f1f5ff01dcd0")
swish Ran yongchao97/diffusion_inversion/src/models/efficientnet.py
code served (permissive licence) · get_code("8737c82de631cffc")
center_crop Not yet run yongchao97/diffusion_inversion/src/dataset.py
code served (permissive licence) · get_code("7be19549d8b0d6dc")
configure_dataloader Not yet run yongchao97/diffusion_inversion/src/dataset.py
code served (permissive licence) · get_code("da8e9de994efee01")
get_full_repo_name Not yet run yongchao97/diffusion_inversion/src/diffuser_inversion.py
code served (permissive licence) · get_code("d7a3942d0d5c6041")
np_tile_imgs Not yet run yongchao97/diffusion_inversion/src/diffuser_inversion.py
code served (permissive licence) · get_code("d59f0a2fa4f321fc")
test Not yet run yongchao97/diffusion_inversion/src/train_net.py
code served (permissive licence) · get_code("94d7f749a6ed65f8")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Acquiring high-quality data for training discriminative models is a crucial yet challenging aspect of building effective predictive systems. In this paper, we present Diffusion Inversion, a simple yet effective method that leverages the pre-trained generative model, Stable Diffusion, to generate diverse, high-quality training data for image classification. Our approach captures the original data distribution and ensures data coverage by inverting images to the latent space of Stable Diffusion, and generates diverse novel training images by conditioning the generative model on noisy versions of these vectors. We identify three key components that allow our generated images to successfully supplant the original dataset, leading to a 2-3x enhancement in sample complexity and a 6.5x decrease in sampling time. Moreover, our approach consistently outperforms generic prompt-based steering methods and KNN retrieval baseline across a wide range of datasets. Additionally, we demonstrate the compatibility of our approach with widely-used data augmentation techniques, as well as the reliability of the generated data in supporting various neural architectures and enhancing few-shot learning.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2305.15316")
get_code_for_paper("2305.15316")
have("2305.15316")

Connect an agent — have() is free.