DeepAI AI Chat
Log In Sign Up

Shift Equivariance in Object Detection

by   Marco Manfredi, et al.

Robustness to small image translations is a highly desirable property for object detectors. However, recent works have shown that CNN-based classifiers are not shift invariant. It is unclear to what extent this could impact object detection, mainly because of the architectural differences between the two and the dimensionality of the prediction space of modern detectors. To assess shift equivariance of object detection models end-to-end, in this paper we propose an evaluation metric, built upon a greedy search of the lower and upper bounds of the mean average precision on a shifted image set. Our new metric shows that modern object detection architectures, no matter if one-stage or two-stage, anchor-based or anchor-free, are sensitive to even one pixel shift to the input images. Furthermore, we investigate several possible solutions to this problem, both taken from the literature and newly proposed, quantifying the effectiveness of each one with the suggested metric. Our results indicate that none of these methods can provide full shift equivariance. Measuring and analyzing the extent of shift variance of different models and the contributions of possible factors, is a first step towards being able to devise methods that mitigate or even leverage such variabilities.


page 2

page 7

page 10


IENet: Interacting Embranchment One Stage Anchor Free Detector for Orientation Aerial Object Detection

Object detection in aerial images is a challenging task due to its lack ...

Towards Improving Calibration in Object Detection Under Domain Shift

The increasing use of deep neural networks in safety-critical applicatio...

Adapted Center and Scale Prediction: More Stable and More Accurate

Pedestrian detection benefits from deep learning technology and gains ra...

Analysing domain shift factors between videos and images for object detection

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

Shift Variance in Scene Text Detection

Theory of convolutional neural networks suggests the property of shift e...

Alias-Free Convnets: Fractional Shift Invariance via Polynomial Activations

Although CNNs are believed to be invariant to translations, recent works...

Learning to Identify Drilling Defects in Turbine Blades with Single Stage Detectors

Nondestructive testing (NDT) is widely applied to defect identification ...