Peer Learning for Unbiased Scene Graph Generation

TL;DR

This paper introduces a peer learning framework with predicate sampling and consensus voting to improve unbiased scene graph generation, effectively handling predicate imbalance and achieving state-of-the-art results.

Contribution

It proposes a novel peer learning approach with predicate sampling and consensus voting to address predicate imbalance in scene graph generation.

Findings

Outperforms previous methods on Visual Genome

Achieves new state-of-the-art on SGCls task with 31.6 mean

Demonstrates effectiveness of peer learning in USGG

Abstract

Unbiased scene graph generation (USGG) is a challenging task that requires predicting diverse and heavily imbalanced predicates between objects in an image. To address this, we propose a novel framework peer learning that uses predicate sampling and consensus voting (PSCV) to encourage multiple peers to learn from each other. Predicate sampling divides the predicate classes into sub-distributions based on frequency, and assigns different peers to handle each sub-distribution or combinations of them. Consensus voting ensembles the peers' complementary predicate knowledge by emphasizing majority opinion and diminishing minority opinion. Experiments on Visual Genome show that PSCV outperforms previous methods and achieves a new state-of-the-art on SGCls task with 31.6 mean.