Comparison of Cross-Entropy, Dice, and Focal Loss for Sea Ice Type Segmentation

• URL: http://arxiv.org/abs/2310.17135v1
• Date: Thu, 26 Oct 2023 04:18:00 GMT
• Title: Comparison of Cross-Entropy, Dice, and Focal Loss for Sea Ice Type Segmentation
• Authors: Rafael Pires de Lima, Behzad Vahedi, Morteza Karimzadeh
• Abstract summary: We show how three loss functions affect the performance of CNN models trained to predict the dominant ice type in Sentinel-1 images. Despite the fact that Dice and Focal loss produce higher metrics, results from cross-entropy seem generally more physically consistent.
• Score: 1.4364491422470593
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Up-to-date sea ice charts are crucial for safer navigation in ice-infested waters. Recently, Convolutional Neural Network (CNN) models show the potential to accelerate the generation of ice maps for large regions. However, results from CNN models still need to undergo scrutiny as higher metrics performance not always translate to adequate outputs. Sea ice type classes are imbalanced, requiring special treatment during training. We evaluate how three different loss functions, some developed for imbalanced class problems, affect the performance of CNN models trained to predict the dominant ice type in Sentinel-1 images. Despite the fact that Dice and Focal loss produce higher metrics, results from cross-entropy seem generally more physically consistent.

Related papers

• Graph Neural Networks for Emulation of Finite-Element Ice Dynamics in Greenland and Antarctic Ice Sheets [0.0]
  equivariant graph convolutional network (EGCN) is an emulator for the ice sheet dynamics modeling. EGCN reproduces ice thickness and velocity changes in the Helheim Glacier, Greenland, and Pine Island Glacier, Antarctica, with 260 times and 44 times faster computation time, respectively.
  arXiv  Detail & Related papers  (2024-06-26T15:18:49Z)
• Partial Label Learning with Focal Loss for Sea Ice Classification Based on Ice Charts [2.0270474485799017]
  We present a novel GeoAI approach to training sea ice classification by formalizing it as a partial label learning task with explicit confidence scores. We treat the polygon-level labels as candidate partial labels, assign the corresponding ice concentrations as confidence scores to each label, and integrate them with focal loss to train a Convolutional Neural Network (CNN) Our proposed approach leads to enhanced performance for sea ice classification in Sentinel-1 dual-polarized SAR images, improving classification accuracy (from 87% to 92%) and weighted average F-1 score (from 90% to 93%) compared to the conventional training approach.
  arXiv  Detail & Related papers  (2024-06-05T22:49:30Z)
• Deep Learning on SAR Imagery: Transfer Learning Versus Randomly Initialized Weights [1.6969743728555278]
  We compare the performance of deep learning models trained from scratch and pretrained models that we fine-tune for this purpose. Our results show that pre-trained models lead to better results, especially on test samples from the melt season.
  arXiv  Detail & Related papers  (2023-10-26T03:52:54Z)
• Boosted Dynamic Neural Networks [53.559833501288146]
  A typical EDNN has multiple prediction heads at different layers of the network backbone. To optimize the model, these prediction heads together with the network backbone are trained on every batch of training data. Treating training and testing inputs differently at the two phases will cause the mismatch between training and testing data distributions. We formulate an EDNN as an additive model inspired by gradient boosting, and propose multiple training techniques to optimize the model effectively.
  arXiv  Detail & Related papers  (2022-11-30T04:23:12Z)
• Detecting train driveshaft damages using accelerometer signals and Differential Convolutional Neural Networks [67.60224656603823]
  This paper proposes the development of a railway axle condition monitoring system based on advanced 2D-Convolutional Neural Network (CNN) architectures. The resultant system converts the railway axle vibration signals into time-frequency domain representations, i.e., spectrograms, and, thus, trains a two-dimensional CNN to classify them depending on their cracks.
  arXiv  Detail & Related papers  (2022-11-15T15:04:06Z)
• Network Augmentation for Tiny Deep Learning [73.57192520534585]
  We introduce Network Augmentation (NetAug), a new training method for improving the performance of tiny neural networks. We demonstrate the effectiveness of NetAug on image classification and object detection.
  arXiv  Detail & Related papers  (2021-10-17T18:48:41Z)
• Churn Reduction via Distillation [54.5952282395487]
  We show an equivalence between training with distillation using the base model as the teacher and training with an explicit constraint on the predictive churn. We then show that distillation performs strongly for low churn training against a number of recent baselines.
  arXiv  Detail & Related papers  (2021-06-04T18:03:31Z)
• BreakingBED -- Breaking Binary and Efficient Deep Neural Networks by Adversarial Attacks [65.2021953284622]
  We study robustness of CNNs against white-box and black-box adversarial attacks. Results are shown for distilled CNNs, agent-based state-of-the-art pruned models, and binarized neural networks.
  arXiv  Detail & Related papers  (2021-03-14T20:43:19Z)
• On Mathews Correlation Coefficient and Improved Distance Map Loss for Automatic Glacier Calving Front Segmentation in SAR Imagery [7.64750171496838]
  The vast majority of outlet glaciers and ice streams of the polar ice sheets end in the ocean. Ice mass loss via calving of the glaciers into the ocean has increased over the last few decades. Deep neural network-based semantic segmentation pipelines can be used to delineate the acquired SAR imagery.
  arXiv  Detail & Related papers  (2021-02-16T17:53:34Z)
• Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effect [95.37587481952487]
  Long-tailed classification is the key to deep learning at scale. Existing methods are mainly based on re-weighting/resamplings that lack a fundamental theory. In this paper, we establish a causal inference framework, which not only unravels the whys of previous methods, but also derives a new principled solution.
  arXiv  Detail & Related papers  (2020-09-28T00:32:11Z)
• Using neural networks to predict icephobic performance [0.5076419064097732]
  Icephobic surfaces inspired by superhydrophobic surfaces offer a passive solution to the problem of icing. Modeling icephobicity is challenging because some material features that aid superhydrophobicity can adversely affect the icephobic performance. This study presents a new approach based on artificial neural networks to model icephobicity.
  arXiv  Detail & Related papers  (2020-07-31T05:37:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.