# Optimizing the F-measure for Threshold-free Salient Object Detection

arXiv: https://arxiv.org/abs/1805.07567 · HTML: https://mlsanity.com/paper/1805.07567/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 May 2018.

## Built on by (most cited first, 15 of 15)

- [Salient Object Detection via Integrity Learning](https://mlsanity.com/paper/2101.07663/) (2021)
- [Boundary-Aware Segmentation Network for Mobile and Web Applications](https://mlsanity.com/paper/2101.04704/) (2021)
- [Revisiting Image Pyramid Structure for High Resolution Salient Object Detection](https://mlsanity.com/paper/2209.09475/) (2022)
- [Disentangled High Quality Salient Object Detection](https://mlsanity.com/paper/2108.03551/) (2021)
- [Group Collaborative Learning for Co-Salient Object Detection](https://mlsanity.com/paper/2104.01108/) (2021)
- [BiconNet: An Edge-preserved Connectivity-based Approach for Salient Object Detection](https://mlsanity.com/paper/2103.00334/) (2021)
- [Progressive Self-Guided Loss for Salient Object Detection](https://mlsanity.com/paper/2101.02412/) (2021)
- [CoSformer: Detecting Co-Salient Object with Transformers](https://mlsanity.com/paper/2104.14729/) (2021)
- [Cascade Network with Guided Loss and Hybrid Attention for Finding Good Correspondences](https://mlsanity.com/paper/2102.00411/) (2021)
- [Benchmarking Deep Models for Salient Object Detection](https://mlsanity.com/paper/2202.02925/) (2022)
- [Disentangle Saliency Detection into Cascaded Detail Modeling and Body Filling](https://mlsanity.com/paper/2202.04112/) (2022)
- [Spatial Coherence Loss: All Objects Matter in Salient and Camouflaged Object Detection](https://mlsanity.com/paper/2402.18698/) (2024)
- [RDNet: Region Proportion-Aware Dynamic Adaptive Salient Object Detection Network in Optical Remote Sensing Images](https://mlsanity.com/paper/2603.12215/) (2026)
- [G2HFNet: GeoGran-Aware Hierarchical Feature Fusion Network for Salient Object Detection in Optical Remote Sensing Images](https://mlsanity.com/paper/2603.12680/) (2026)
- [RSONet: Region-guided Selective Optimization Network for RGB-T Salient Object Detection](https://mlsanity.com/paper/2603.12685/) (2026)

## Abstract (arXiv)

Current CNN-based solutions to salient object detection (SOD) mainly rely on the optimization of cross-entropy loss (CELoss). Then the quality of detected saliency maps is often evaluated in terms of F-measure. In this paper, we investigate an interesting issue: can we consistently use the F-measure formulation in both training and evaluation for SOD? By reformulating the standard F-measure we propose the relaxed F-measure which is differentiable w.r.t the posterior and can be easily appended to the back of CNNs as the loss function. Compared to the conventional cross-entropy loss of which the gradients decrease dramatically in the saturated area, our loss function, named FLoss, holds considerable gradients even when the activation approaches the target. Consequently, the FLoss can continuously force the network to produce polarized activations. Comprehensive benchmarks on several popular datasets show that FLoss outperforms the state-of-the-art with a considerable margin. More specifically, due to the polarized predictions, our method is able to obtain high-quality saliency maps without carefully tuning the optimal threshold, showing significant advantages in real-world applications.