OmniLabel

Emerging

1papers using it

2026first seen

The 'OmniLabel' dataset/benchmark contains a diverse set of images and annotations used to evaluate the robustness of open-vocabulary object detection models in recognizing objects across varying backgrounds and environmental conditions.

🔎 Find this dataset

Papers using OmniLabel (1)

Consistency Beyond Contrast: Enhancing Open-Vocabulary Object Detection Robustness via Contextual Consistency Learning2026