Recovering the Unbiased Scene Graphs from the Biased Ones

by   Meng-Jiun Chiou, et al.

Given input images, scene graph generation (SGG) aims to produce comprehensive, graphical representations describing visual relationships among salient objects. Recently, more efforts have been paid to the long tail problem in SGG; however, the imbalance in the fraction of missing labels of different classes, or reporting bias, exacerbating the long tail is rarely considered and cannot be solved by the existing debiasing methods. In this paper we show that, due to the missing labels, SGG can be viewed as a "Learning from Positive and Unlabeled data" (PU learning) problem, where the reporting bias can be removed by recovering the unbiased probabilities from the biased ones by utilizing label frequencies, i.e., the per-class fraction of labeled, positive examples in all the positive examples. To obtain accurate label frequency estimates, we propose Dynamic Label Frequency Estimation (DLFE) to take advantage of training-time data augmentation and average over multiple training iterations to introduce more valid examples. Extensive experiments show that DLFE is more effective in estimating label frequencies than a naive variant of the traditional estimate, and DLFE significantly alleviates the long tail and achieves state-of-the-art debiasing performance on the VG dataset. We also show qualitatively that SGG models with DLFE produce prominently more balanced and unbiased scene graphs.


page 3

page 7

page 8

page 12

page 17


HiLo: Exploiting High Low Frequency Relations for Unbiased Panoptic Scene Graph Generation

Panoptic Scene Graph generation (PSG) is a recently proposed task in ima...

Triple Correlations-Guided Label Supplementation for Unbiased Video Scene Graph Generation

Video-based scene graph generation (VidSGG) is an approach that aims to ...

Taking A Closer Look at Visual Relation: Unbiased Video Scene Graph Generation with Decoupled Label Learning

Current video-based scene graph generation (VidSGG) methods have been fo...

CogTree: Cognition Tree Loss for Unbiased Scene Graph Generation

Scene graphs are semantic abstraction of images that encourage visual un...

PCPL: Predicate-Correlation Perception Learning for Unbiased Scene Graph Generation

Today, scene graph generation(SGG) task is largely limited in realistic ...

Unbiased Scene Graph Generation using Predicate Similarities

Scene Graphs are widely applied in computer vision as a graphical repres...

Vision Relation Transformer for Unbiased Scene Graph Generation

Recent years have seen a growing interest in Scene Graph Generation (SGG...

Please sign up or login with your details

Forgot password? Click here to reset