Amulet: Aggregating Multi-level Convolutional Features for Salient Object Detection

TL;DR

Amulet introduces a flexible framework that effectively combines multi-level convolutional features to improve salient object detection accuracy, integrating semantic and detail information for better boundary and object labeling.

Contribution

This work proposes a novel multi-level feature aggregation framework for salient object detection, enhancing boundary accuracy and semantic understanding compared to existing methods.

Findings

Outperforms state-of-the-art methods on multiple evaluation metrics

Effectively integrates multi-resolution features for improved saliency maps

Achieves accurate boundary inference and semantic enhancement

Abstract

Fully convolutional neural networks (FCNs) have shown outstanding performance in many dense labeling problems. One key pillar of these successes is mining relevant information from features in convolutional layers. However, how to better aggregate multi-level convolutional feature maps for salient object detection is underexplored. In this work, we present Amulet, a generic aggregating multi-level convolutional feature framework for salient object detection. Our framework first integrates multi-level feature maps into multiple resolutions, which simultaneously incorporate coarse semantics and fine details. Then it adaptively learns to combine these feature maps at each resolution and predict saliency maps with the combined features. Finally, the predicted results are efficiently fused to generate the final saliency map. In addition, to achieve accurate boundary inference and semantic enhancement, edge-aware feature maps in low-level layers and the predicted results of low resolution features are recursively embedded into the learning framework. By aggregating multi-level convolutional features in this efficient and flexible manner, the proposed saliency model provides accurate salient object labeling. Comprehensive experiments demonstrate that our method performs favorably against state-of-the art approaches in terms of near all compared evaluation metrics.