Hallucination In Object Detection -- A Study In Visual Part Verification
This addresses a reliability issue in object detection for applications like quality control or safety checks, though it is incremental as it focuses on a specific dataset and task.
The paper tackles the problem of object detectors hallucinating missing objects, which is critical for visual part verification tasks, and introduces the DelftBikes dataset with 10,000 bike images and 22 annotated parts per image to evaluate this issue, showing that popular detectors perform poorly in this context.
We show that object detectors can hallucinate and detect missing objects; potentially even accurately localized at their expected, but non-existing, position. This is particularly problematic for applications that rely on visual part verification: detecting if an object part is present or absent. We show how popular object detectors hallucinate objects in a visual part verification task and introduce the first visual part verification dataset: DelftBikes, which has 10,000 bike photographs, with 22 densely annotated parts per image, where some parts may be missing. We explicitly annotated an extra object state label for each part to reflect if a part is missing or intact. We propose to evaluate visual part verification by relying on recall and compare popular object detectors on DelftBikes.