A survey of loss functions for semantic segmentation

Abstract

Image Segmentation has been an active field of research as it has a wide range of applications, ranging from automated disease detection to self-driving cars. In the past five years, various papers came up with different objective loss functions used in different cases such as biased data, sparse segmentation, etc. In this paper, we have summarized some of the well-known loss functions widely used for Image Segmentation and listed out the cases where their usage can help in fast and better convergence of a model. Furthermore, we have also introduced a new log-cosh dice loss function and compared its performance on the NBFS skull-segmentation open-source data-set with widely used loss functions. We also showcased that certain loss functions perform well across all data-sets and can be taken as a good baseline choice in unknown data distribution scenarios. Our code is available at Github: https://github.com/shruti-jadon/Semantic-Segmentation-Loss-Functions.

arXiv comment5 pages, 5 figures, 2 tables

References in corpus (5)

• Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations
• The Real-World-Weight Cross-Entropy Loss Function: Modeling the Costs of Mislabeling
• Asymmetric Loss Functions and Deep Densely Connected Networks for Highly Imbalanced Medical Image Segmentation: Application to Multiple Sclerosis Lesion Detection
• 3D Segmentation with Exponential Logarithmic Loss for Highly Unbalanced Object Sizes
• A comparative study of 2D image segmentation algorithms for traumatic brain lesions using CT data from the ProTECTIII multicenter clinical trial

Cited by in corpus (29)

• A Survey on Deep Learning-based Architectures for Semantic Segmentation on 2D images
• Deep learning for brain metastasis detection and segmentation in longitudinal MRI data
• Development of Skip Connection in Deep Neural Networks for Computer Vision and Medical Image Analysis: A Survey
• Graph Neural Networks for Low-Energy Event Classification & Reconstruction in IceCube
• Degradation-invariant Enhancement of Fundus Images via Pyramid Constraint Network
• RLAS-BIABC: A Reinforcement Learning-Based Answer Selection Using the BERT Model Boosted by an Improved ABC Algorithm
• Lebanon Solar Rooftop Potential Assessment using Buildings Segmentation from Aerial Images
• Machine-learning recovery of foreground wedge-removed 21-cm light cones for high-$z$ galaxy mapping
• Multiclass Segmentation using Teeth Attention Modules for Dental X-ray Images
• Grasping as Inference: Reactive Grasping in Heavily Cluttered Environment
• Robust Lane Detection through Self Pre-training with Masked Sequential Autoencoders and Fine-tuning with Customized PolyLoss
• Automatic Detection of Interplanetary Coronal Mass Ejections in Solar Wind In Situ Data
• Deformation Driven Seq2Seq Longitudinal Tumor and Organs-at-Risk Prediction for Radiotherapy
• GenSelfDiff-HIS: Generative Self-Supervision Using Diffusion for Histopathological Image Segmentation
• Evaluation of Data Augmentation and Loss Functions in Semantic Image Segmentation for Drilling Tool Wear Detection
• Training a Distributed Acoustic Sensing Traffic Monitoring Network With Video Inputs
• Image segmentation of treated and untreated tumor spheroids by Fully Convolutional Networks
• The Dice loss in the context of missing or empty labels: Introducing $\Phi$ and $\epsilon$
• A Simple and Robust Framework for Cross-Modality Medical Image Segmentation applied to Vision Transformers
• Majority or Minority: Data Imbalance Learning Method for Named Entity Recognition
• Deep Radar Inverse Sensor Models for Dynamic Occupancy Grid Maps
• Datacube segmentation via Deep Spectral Clustering
• Slim U-Net: Efficient Anatomical Feature Preserving U-net Architecture for Ultrasound Image Segmentation
• AI Powered Road Network Prediction with Multi-Modal Data
• Advancing biological super-resolution microscopy through deep learning: a brief review
• On Mathews Correlation Coefficient and Improved Distance Map Loss for Automatic Glacier Calving Front Segmentation in SAR Imagery
• Finding NeMO: A Geometry-Aware Representation of Template Views for Few-Shot Perception
• Automated Mosaic Tesserae Segmentation via Deep Learning Techniques
• DDU-Net: A Domain Decomposition-Based CNN for High-Resolution Image Segmentation on Multiple GPUs