PulseAugur
LIVE 12:28:20
research · [3 sources] ·
0
research

New pipeline generates synthetic industrial defects to augment scarce real-world data

By PulseAugur Editorial · Summary by gemini-2.5-flash-lite from 3 sources

Researchers have developed SynSur, an end-to-end pipeline for generating synthetic industrial surface defects to address the scarcity of labeled data in defect detection. The pipeline combines vision-language models, LoRA-adapted diffusion, and mask-guided inpainting to create realistic defect samples. Experiments show that while synthetic data alone cannot replace real data, it can enhance performance when combined with existing datasets, particularly in improving training regimes and cross-domain transfer. AI

Summary written by gemini-2.5-flash-lite from 3 sources. How we write summaries →

IMPACT Enhances industrial defect detection by augmenting scarce real-world datasets with realistic synthetic samples.

RANK_REASON The cluster describes an academic paper detailing a new method for synthetic data generation.

Read on arXiv cs.CV →

COVERAGE [3]

  1. Hugging Face Daily Papers TIER_1 ·

    SynSur: An end-to-end generative pipeline for synthetic industrial surface defect generation and detection

    The bottleneck in learning-based industrial defect detection is often limited not by model capacity, but by the scarcity of labeled defect data: defects are rare, annotations are expensive, and collecting balanced training sets is slow. We present an end-to-end pipeline for synth…

  2. arXiv cs.CV TIER_1 · Paul Julius K\"uhn, Mika Pommeranz, Arjan Kuijper, Saptarshi Neil Sinha ·

    SynSur: An end-to-end generative pipeline for synthetic industrial surface defect generation and detection

    arXiv:2604.26633v1 Announce Type: new Abstract: The bottleneck in learning-based industrial defect detection is often limited not by model capacity, but by the scarcity of labeled defect data: defects are rare, annotations are expensive, and collecting balanced training sets is s…

  3. arXiv cs.CV TIER_1 · Saptarshi Neil Sinha ·

    SynSur: An end-to-end generative pipeline for synthetic industrial surface defect generation and detection

    The bottleneck in learning-based industrial defect detection is often limited not by model capacity, but by the scarcity of labeled defect data: defects are rare, annotations are expensive, and collecting balanced training sets is slow. We present an end-to-end pipeline for synth…

RELATED ENTITIES

RELATED TOPICS