Application of Maximum Likelihood Decision Rules for Handling Class Imbalance in Semantic Segmentation

Chan, Robin; Rottmann, Matthias; Gottschalk, Hanno; Hüger, Fabian; Schlicht, Peter

doi:10.3850/978-981-14-8593-0_5748-cd

Cited by 22 publications

(24 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The stream datapoints that enter are decided by AL, whether or not to perform the datapoint annotation (as it arrives). In stream-based AL, the annotation process is a discrete action, and Q-learning [37] is the preferred RL technique [20], [21]. Pool-based AL, on the other hand, is concerned with all the possible annotated datapoints, and it is characterized naturally by using continuous vectors, making it impossible for Q-learning.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

A Reinforced Active Learning Algorithm for Semantic Segmentation in Complex Imaging

Ahmadusmani¹,

Watada

Jaafar

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Semantic segmentation annotation helps train computer vision based Artificial Intelligence models where each image pixel is assigned to a specific object class. The model developers try to identify the features helpful for determining the objects of interest by using various supervised deep learning techniques. However, this is a difficult task due to the complexity of object structures. Two difficulties arise in the current approaches for semantic segmentation. The pixel-wise label approach is costly to obtain and is time consuming. Second, the datasets taken for the semantic segmentation task are not balanced since certain classes are present more than the others. This biases the model performance to the most represented ones. We propose a new reinforced active learning strategy based on a deep reinforcement learning algorithm. This work presents a modified Deep Q Learning formulation for active learning. An agent learns the strategy of selecting a subset of small image regions, which are more knowledgeable than the whole set of images from an unlabeled data pool. The decision on the area of selection is dependent on the assumptions and segmentation model uncertainties taken for training purposes. We use the CamVid and RGB indoor test scenes dataset to evaluate the proof of concept. Our results infer that our approach demands more labels from under-represented groups than the baselines, thus enhancing their efficiency and mitigating the class imbalance. Our method's performance is superior to the conventional deep learning models in detecting 8 out of 11 classes on the Camvid road segmentation scene dataset. It achieves an accuracy of 90.56%, a mIoU score of 87.17%, and a BF score of 93.14%. On the SUNRGB indoor scenes dataset, it gives an accuracy of around 75.82% and a BF score of 77.25%, thus outperforming the current state-of-the-art methods.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…Because every step updates the segmentation network and computes the rewards, thus making the task inefficient. In this work, we show how to train an AL model for semantic segmentation using Reinforcement Learning (RL) by maximization of the performance metric, mean Intersection over Union (mIoU) [20], [21].…”

Section: Introductionmentioning

confidence: 99%

A Reinforced Active Learning Algorithm for Semantic Segmentation in Complex Imaging

Ahmadusmani¹,

Watada

Jaafar

et al. 2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Certain strategies can be adopted to improve the model's robustness for D i classes. For example, modifying the class weights in the loss function (Eigen & Fergus, 2015) and adjusting the decision rule (Chan, Rottmann, Hüger, Schlicht, & Gottschalk, 2019) are two techniques that alleviate the effects of class imbalance. These approaches tend to increase the mean class accuracy ( MCA ) while global accuracy ( GA ) may be compromised.…”

Section: Dual Inference Strategymentioning

confidence: 99%

Dual Bayesian inference for risk‐informed vibration‐based damage diagnosis

Sajedi

Liang

2020

Computer aided Civil Eng

View full text Add to dashboard Cite

Automation in structural health monitoring (SHM) has greatly benefited from computer science's recent advances. Unlike images, the existing datasets for other types of input, such as vibration‐based damage data, are relatively smaller, less diverse, and highly imbalanced. Therefore, the reliability of data‐driven models developed for safety‐critical assessments can be questionable. This paper proposes a dual Bayesian inference where damage predictions are accompanied by measuring the model's confidence in predictions. First, it is shown how dual classification is integrated with Bayesian inference. Later, we introduce a surrogate deep learning module to transform the raw uncertainty output into an easily interpretable prediction uncertainty index (PUI). The PUI metric can be used to alarm a decision‐maker of the potential mistakes. The proposed dual Bayesian models are investigated on a 2D structure with seven different sensor layouts. Our approach yields increased robustness for different metrics compared with the benchmark. In addition to the performance boost, PUI information paves the way for a risk‐informed implementation of deep learning models in vibration‐based damage diagnosis.

show abstract

“…As alternatives we propose a decision principle -the maximum likelihood (ML) decision rule [20] -that looks out for the best fit of the data to a given semantic class. We review the false negative detection [21] using the ML decision rule for semantic segmentation and also discuss cost based decision rules in general along with the problems of setting the cost structure up [22].…”

Section: Introductionmentioning

confidence: 99%

Detection of False Positive and False Negative Samples in Semantic Segmentation

Rottmann¹,

Maag²,

Chan³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

In recent years, deep learning methods have outperformed other methods in image recognition. This has fostered imagination of potential application of deep learning technology including safety relevant applications like the interpretation of medical images or autonomous driving. The passage from assistance of a human decision maker to ever more automated systems however increases the need to properly handle the failure modes of deep learning modules. In this contribution, we review a set of techniques for the self-monitoring of machine-learning algorithms based on uncertainty quantification. In particular, we apply this to the task of semantic segmentation, where the machine learning algorithm decomposes an image according to semantic categories. We discuss false positive and false negative error modes at instance-level and review techniques for the detection of such errors that have been recently proposed by the authors. We also give an outlook on future research directions.

show abstract

Application of Maximum Likelihood Decision Rules for Handling Class Imbalance in Semantic Segmentation

Cited by 22 publications

References 0 publications

A Reinforced Active Learning Algorithm for Semantic Segmentation in Complex Imaging

A Reinforced Active Learning Algorithm for Semantic Segmentation in Complex Imaging

Dual Bayesian inference for risk‐informed vibration‐based damage diagnosis

Detection of False Positive and False Negative Samples in Semantic Segmentation

Contact Info

Product

Resources

About