Hybrid Knowledge Routed Modules for Large-scale Object Detection

TL;DR

This paper introduces Hybrid Knowledge Routed Modules (HKRM) that leverage explicit and implicit human knowledge to improve large-scale object detection, especially in long-tail and confusing categories, by enabling semantic reasoning.

Contribution

The paper proposes HKRM, a novel module that integrates structured and implicit knowledge for enhanced reasoning in object detection networks.

Findings

Achieved 34.5% mAP improvement on VisualGenome

Achieved 30.4% mAP improvement on ADE

Demonstrated effectiveness in large-scale, long-tail scenarios

Abstract

The dominant object detection approaches treat the recognition of each region separately and overlook crucial semantic correlations between objects in one scene. This paradigm leads to substantial performance drop when facing heavy long-tail problems, where very few samples are available for rare classes and plenty of confusing categories exists. We exploit diverse human commonsense knowledge for reasoning over large-scale object categories and reaching semantic coherency within one image. Particularly, we present Hybrid Knowledge Routed Modules (HKRM) that incorporates the reasoning routed by two kinds of knowledge forms: an explicit knowledge module for structured constraints that are summarized with linguistic knowledge (e.g. shared attributes, relationships) about concepts; and an implicit knowledge module that depicts some implicit constraints (e.g. common spatial layouts). By functioning over a region-to-region graph, both modules can be individualized and adapted to coordinate with visual patterns in each image, guided by specific knowledge forms. HKRM are light-weight, general-purpose and extensible by easily incorporating multiple knowledge to endow any detection networks the ability of global semantic reasoning. Experiments on large-scale object detection benchmarks show HKRM obtains around 34.5% improvement on VisualGenome (1000 categories) and 30.4% on ADE in terms of mAP. Codes and trained model can be found in https://github.com/chanyn/HKRM.