Hierarchical Context Embedding for Region-based Object Detection

TL;DR

This paper introduces a Hierarchical Context Embedding framework that enhances region-based object detectors by incorporating multi-level contextual cues, significantly improving classification accuracy across various models.

Contribution

The paper proposes a novel hierarchical context embedding method with image-level and region-level cues, along with fusion strategies, to improve object detection performance.

Findings

Significant improvements on FPN, Cascade R-CNN, and Mask R-CNN.

Effective integration of hierarchical context boosts detection accuracy.

Framework is flexible and can be applied to various detectors.

Abstract

State-of-the-art two-stage object detectors apply a classifier to a sparse set of object proposals, relying on region-wise features extracted by RoIPool or RoIAlign as inputs. The region-wise features, in spite of aligning well with the proposal locations, may still lack the crucial context information which is necessary for filtering out noisy background detections, as well as recognizing objects possessing no distinctive appearances. To address this issue, we present a simple but effective Hierarchical Context Embedding (HCE) framework, which can be applied as a plug-and-play component, to facilitate the classification ability of a series of region-based detectors by mining contextual cues. Specifically, to advance the recognition of context-dependent object categories, we propose an image-level categorical embedding module which leverages the holistic image-level context to learn object-level concepts. Then, novel RoI features are generated by exploiting hierarchically embedded context information beneath both whole images and interested regions, which are also complementary to conventional RoI features. Moreover, to make full use of our hierarchical contextual RoI features, we propose the early-and-late fusion strategies (i.e., feature fusion and confidence fusion), which can be combined to boost the classification accuracy of region-based detectors. Comprehensive experiments demonstrate that our HCE framework is flexible and generalizable, leading to significant and consistent improvements upon various region-based detectors, including FPN, Cascade R-CNN and Mask R-CNN.