REVIEW 4 cited by
Training on Thin Air: Improve Image Classification with Generated Data
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original 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.
Forward citations
Cited by 4 Pith papers
-
SoK: Can Synthetic Images Replace Real Data? A Survey of Utility and Privacy of Synthetic Image Generation
A systematic survey and benchmark showing that diffusion-based synthetic data can achieve better utility-privacy tradeoffs than DP-SGD on real data for some image classifiers, with the best release strategy depending ...
-
Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting
Post-generation selection via Homogeneous-Heterogeneous real-data splits and a fidelity-diversity score raises synthetic-image utility for classification and segmentation without retraining generators.
-
Beyond Objects: Contextual Synthetic Data Generation for Fine-Grained Classification
Conditioning a fine-tuned text-to-image model on per-image background/pose captions and then randomly recombining those contexts across classes improves few-shot fine-grained classifier accuracy.
-
Generate, Refine, and Encode: Leveraging Synthesized Novel Samples for On-the-Fly Fine-Grained Category Discovery
DiffGRE generates synthetic images through cross-image interpolation in diffusion and CLIP latent spaces, filters them for diversity, and uses pseudo-labels to improve on-the-fly fine-grained category discovery.
Discussion (0). Sign in to comment.