<- Back to Datasets

Natural Adversarial Object (NAO)

A dataset to evaluate robustness of object detection models

Natural Adversarial Object (NAO)

We introduce a new dataset, Natural Adversarial Objects (NAO), to evaluate the robustness of object detection models. NAO contains 7,934 images and 9,943 objects that are unmodified and representative of real-world scenarios, but cause state-of-the-art detection models to misclassify with high confidence. The mean average precision (mAP) of EfficientDet-D7 drops 74.5% when evaluated on NAO compared to the standard MSCOCO validation set.

View this Dataset
->
Scale AI, ML Collective and Allen Institute for AI, Masakhane
View author website
Task
Object Detection
Annotation Types
Bounding Boxes
7934
Items
Classes
7934
Labels
Models using this dataset
Last updated on 
January 20, 2022
Licensed under 
Unknown
Gain control of your training data
15,000+ ML engineers can’t be wrong