TRACER: Extreme Attention Guided Salient Object Tracing Network

14 Dec 2021  ·  Min Seok Lee, WooSeok Shin, Sung Won Han ·

Existing studies on salient object detection (SOD) focus on extracting distinct objects with edge information and aggregating multi-level features to improve SOD performance. To achieve satisfactory performance, the methods employ refined edge information and low multi-level discrepancy. However, both performance gain and computational efficiency cannot be attained, which has motivated us to study the inefficiencies in existing encoder-decoder structures to avoid this trade-off. We propose TRACER, which detects salient objects with explicit edges by incorporating attention guided tracing modules. We employ a masked edge attention module at the end of the first encoder using a fast Fourier transform to propagate the refined edge information to the downstream feature extraction. In the multi-level aggregation phase, the union attention module identifies the complementary channel and important spatial information. To improve the decoder performance and computational efficiency, we minimize the decoder block usage with object attention module. This module extracts undetected objects and edge information from refined channels and spatial representations. Subsequently, we propose an adaptive pixel intensity loss function to deal with the relatively important pixels unlike conventional loss functions which treat all pixels equally. A comparison with 13 existing methods reveals that TRACER achieves state-of-the-art performance on five benchmark datasets. We have released TRACER at

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
RGB Salient Object Detection DUT-OMRON TRACER-(ResNet50) MAE 0.050 # 4
RGB Salient Object Detection DUT-OMRON TRACER-TE7 MAE 0.045 # 1
F-measure 0.849 # 1
S-Measure 0.855 # 2
mean F-Measure 0.798 # 1
RGB Salient Object Detection DUTS-TE TRACER-(ResNet50) MAE 0.035 # 8
RGB Salient Object Detection DUTS-TE TRACER-TE7 MAE 0.022 # 1
F-measure 0.932 # 1
S-Measure 0.919 # 2
mean F-Measure 0.904 # 1
RGB Salient Object Detection ECSSD TRACER-(ResNet50) MAE 0.033 # 5
RGB Salient Object Detection ECSSD TRACER-TE7 MAE 0.026 # 2
F-measure 0.961 # 1
S-Measure 0.935 # 3
mean F-Measure 0.940 # 1
RGB Salient Object Detection HKU-IS TRACER-(ResNet50) MAE 0.028 # 6
RGB Salient Object Detection HKU-IS TRACER-TE7 MAE 0.020 # 1
F-measure 0.954 # 2
S-Measure 0.932 # 2
mean F-Measure 0.934 # 1
RGB Salient Object Detection PASCAL-S TRACER-TE7 MAE 0.047 # 1
F-measure 0.909 # 1
mean F-Measure 0.874 # 1
S-Measure 0.882 # 2


No methods listed for this paper. Add relevant methods here