Universal Noise Annotation: Unveiling the Impact of Noisy annotation on Object Detection
This addresses the challenge of noisy labels for object detection practitioners, but it is incremental as it builds on prior work by combining existing noise types into a unified setting.
The paper tackles the problem of noisy annotations in object detection by proposing Universal-Noise Annotation (UNA), a setting that includes all noise types like categorization, localization, missing annotations, and bogus bounding boxes, and analyzes its impact on detector performance, with code and training data shared.
For object detection task with noisy labels, it is important to consider not only categorization noise, as in image classification, but also localization noise, missing annotations, and bogus bounding boxes. However, previous studies have only addressed certain types of noise (e.g., localization or categorization). In this paper, we propose Universal-Noise Annotation (UNA), a more practical setting that encompasses all types of noise that can occur in object detection, and analyze how UNA affects the performance of the detector. We analyzed the development direction of previous works of detection algorithms and examined the factors that impact the robustness of detection model learning method. We open-source the code for injecting UNA into the dataset and all the training log and weight are also shared.