The Elephant in the Room

by   Amir Rosenfeld, et al.

We showcase a family of common failures of state-of-the art object detectors. These are obtained by replacing image sub-regions by another sub-image that contains a trained object. We call this "object transplanting". Modifying an image in this manner is shown to have a non-local impact on object detection. Slight changes in object position can affect its identity according to an object detector as well as that of other objects in the image. We provide some analysis and suggest possible reasons for the reported phenomena.


page 2

page 4

page 5

page 8

page 9

page 10

page 11

page 12


The Sierpinski Object in the Scott Realizability Topos

We study the Sierpinski Object in the Scott Realizability Topos....

Analysing domain shift factors between videos and images for object detection

Object detection is one of the most important challenges in computer vis...

Object 6D Pose Estimation with Non-local Attention

In this paper, we address the challenging task of estimating 6D object p...

Why Accuracy Is Not Enough: The Need for Consistency in Object Detection

Object detectors are vital to many modern computer vision applications. ...

Clustered Object Detection in Aerial Images

Detecting objects in aerial images is challenging for at least two reaso...

Max-Margin Object Detection

Most object detection methods operate by applying a binary classifier to...

Determinantal Point Process as an alternative to NMS

We present a determinantal point process (DPP) inspired alternative to n...

Code Repositories


A simple GUI tool for generating adversarial poses of objects.

view repo