Efficient sampling-based energy function evaluation for ensemble optimization using simulated annealing. (November 2020)

• Record Type: Journal Article
• Authors: Tóth, János; Tomán, Henrietta; Hajdu, András
• Is Part Of: Pattern recognition. Volume 107(2020:Nov.)
• Publication Date: 2020-11
• Subjects: Diabetic retinopathy -- Ensemble -- Microaneurysm detection -- Parameter optimization -- Sampling-based evaluation -- Simulated annealing; 68U10; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2020.107510
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A sampling-based evaluation method is proposed for accelerating optimization with simulated annealing over large image datasets. The sampling strategy is constructed using convergence results for noisy evaluation. The proposed method significantly reduces the time requirement of the optimization while preserves convergence in probability to the global optimum. As an application, an ensemble for diabetic retinopathy pre-screening is created and optimized. Abstract: In this study, we attempted to develop a method for accelerating parameter optimization of an object detector ensemble over large image datasets by using simulated annealing. We propose a novel sampling-based evaluation method that considers the minimum portion of the dataset required in each iteration to maintain solution quality. This approach can be considered a noisy evaluation of the energy. The sample sizes required during the search process are theoretically determined by adapting the convergence results for noisy evaluation. To determine applicability, we prepared and optimized two ensembles for diabetic retinopathy pre-screening based on microaneurysm detection with convolutional neural network-based and traditional object detectors. Our experimental results indicate that the proposed sampling-based evaluation method substantially reduced the computational time required for optimizing the parameters of the ensembles while preserving solution quality.