Ctrl-GenAug: Controllable Generative Augmentation for Medical Sequence Classification

X Zhou, Y Huang, H Dou, S Chen, A Chang… - arxiv preprint arxiv …, 2024 - arxiv.org
In the medical field, the limited availability of large-scale datasets and labor-intensive
annotation processes hinder the performance of deep models. Diffusion-based generative …

Boosting Few-Shot Detection with Large Language Models and Layout-to-Image Synthesis

A Abdullah, N Ebert… - Proceedings of the …, 2024 - openaccess.thecvf.com
Recent advancements in diffusion models have enabled a wide range of works exploiting
their ability to generate high-volume, high-quality data for use in various downstream tasks …

RTGen: Generating Region-Text Pairs for Open-Vocabulary Object Detection

F Chen, H Zhang, Z Yang, H Chen, K Hu… - arxiv preprint arxiv …, 2024 - arxiv.org
Open-vocabulary object detection (OVD) requires solid modeling of the region-semantic
relationship, which could be learned from massive region-text pairs. However, such data is …

No Annotations for Object Detection in Art through Stable Diffusion

P Ramos, N Gonthier, S Khan, Y Nakashima… - arxiv preprint arxiv …, 2024 - arxiv.org
Object detection in art is a valuable tool for the digital humanities, as it allows for faster
identification of objects in artistic and historical images compared to humans. However …

Paint Outside the Box: Synthesizing and Selecting Training Data for Visual Grounding

Z Du, H Li, J Yu, B Li - arxiv preprint arxiv:2412.00684, 2024 - arxiv.org
Visual grounding aims to localize the image regions based on a textual query. Given the
difficulty of large-scale data curation, we investigate how to effectively learn visual grounding …

Sampling Bag of Views for Open-Vocabulary Object Detection

H Choi, J Choe, H Shim - arxiv preprint arxiv:2412.18273, 2024 - arxiv.org
Existing open-vocabulary object detection (OVD) develops methods for testing unseen
categories by aligning object region embeddings with corresponding VLM features. A recent …

OCDet: Object Center Detection via Bounding Box-Aware Heatmap Prediction on Edge Devices with NPUs

C **n, T Motz, A Hartel, E Kasneci - arxiv preprint arxiv:2411.15653, 2024 - arxiv.org
Real-time object localization on edge devices is fundamental for numerous applications,
ranging from surveillance to industrial automation. Traditional frameworks, such as object …

Diverse Generation while Maintaining Semantic Coordination: A Diffusion-Based Data Augmentation Method for Object Detection

S Nie, Z Wang, X Wang, K He - arxiv preprint arxiv:2408.02891, 2024 - arxiv.org
Recent studies emphasize the crucial role of data augmentation in enhancing the
performance of object detection models. However, existing methodologies often struggle to …