Real-Time Salient Object Detection With a Minimum Spanning Tree

In this paper, we present a real-time salient object detection system based on the minimum spanning tree. Due to the fact that background regions are typically connected to the image boundaries, salient objects can be extracted by computing the distances to the boundaries... However, measuring the image boundary connectivity efficiently is a challenging problem. Existing methods either rely on superpixel representation to reduce the processing units or approximate the distance transform. Instead, we propose an exact and iteration free solution on a minimum spanning tree. The minimum spanning tree representation of an image inherently reveals the object geometry information in a scene. Meanwhile, it largely reduces the search space of shortest paths, resulting an efficient and high quality distance transform algorithm. We further introduce a boundary dissimilarity measure to compliment the shortage of distance transform for salient object detection. Extensive evaluations show that the proposed algorithm achieves the leading performance compared to the state-of-the-art methods in terms of efficiency and accuracy. read more

PDF Abstract

Results from the Paper


Ranked #5 on Video Salient Object Detection on MCL (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Benchmark
Video Salient Object Detection DAVIS-2016 MSTM S-Measure 0.566 # 9
MAX E-MEASURE 0.734 # 7
AVERAGE MAE 0.174 # 2
Video Salient Object Detection DAVSOD-Difficult20 MSTM S-Measure 0.488 # 8
max E-measure 0.676 # 5
Average MAE 0.227 # 7
Video Salient Object Detection DAVSOD-easy35 MSTM S-Measure 0.530 # 8
max E-Measure 0.632 # 6
Average MAE 0.214 # 7
Video Salient Object Detection DAVSOD-Normal25 MSTM S-Measure 0.496 # 8
max E-measure 0.573 # 7
Average MAE 0.251 # 7
Video Salient Object Detection FBMS-59 MSTM S-Measure 0.613 # 12
AVERAGE MAE 0.177 # 12
MAX F-MEASURE 0.500 # 12
Video Salient Object Detection MCL MSTM S-Measure 0.700 # 5
MAX E-MEASURE 0.838 # 4
AVERAGE MAE 0.078 # 5
Video Salient Object Detection SegTrack v2 MSTM S-Measure 0.643 # 6
AVERAGE MAE 0.114 # 5
max E-measure 0.733 # 7
Video Salient Object Detection UVSD MSTM S-Measure 0.551 # 7
max E-measure 0.718 # 6
Average MAE 0.145 # 6
Video Salient Object Detection ViSal MSTM S-Measure 0.749 # 5
max E-measure 0.816 # 7
Average MAE 0.095 # 5
Video Salient Object Detection VOS-T MSTM S-Measure 0.657 # 7
max E-measure 0.695 # 7
Average MAE 0.144 # 6

Methods


No methods listed for this paper. Add relevant methods here