MobileSal: Extremely Efficient RGB-D Salient Object Detection

TL;DR

MobileSal is a highly efficient RGB-D salient object detection network that leverages implicit depth restoration and compact pyramid refinement to achieve state-of-the-art performance with minimal computational cost and high speed.

Contribution

The paper introduces MobileSal, a novel mobile network-based RGB-D SOD method with an implicit depth restoration technique and efficient feature aggregation, enabling real-time performance.

Findings

Achieves 450 fps on 320x320 images.

Fewer parameters (6.5M) than competing methods.

Performs favorably on six challenging datasets.

Abstract

The high computational cost of neural networks has prevented recent successes in RGB-D salient object detection (SOD) from benefiting real-world applications. Hence, this paper introduces a novel network, MobileSal, which focuses on efficient RGB-D SOD using mobile networks for deep feature extraction. However, mobile networks are less powerful in feature representation than cumbersome networks. To this end, we observe that the depth information of color images can strengthen the feature representation related to SOD if leveraged properly. Therefore, we propose an implicit depth restoration (IDR) technique to strengthen the mobile networks' feature representation capability for RGB-D SOD. IDR is only adopted in the training phase and is omitted during testing, so it is computationally free. Besides, we propose compact pyramid refinement (CPR) for efficient multi-level feature aggregation to derive salient objects with clear boundaries. With IDR and CPR incorporated, MobileSal performs favorably against state-of-the-art methods on six challenging RGB-D SOD datasets with much faster speed (450fps for the input size of 320 $\times$ 320) and fewer parameters (6.5M). The code is released at https://mmcheng.net/mobilesal.