Correction - Correction YOLO-based intelligent recognition system for hidden dangers at construction sites - Algorithm 1 - Pseudocode

Sep 29, 2026

Correction to a PLOS One open-access article on a YOLO-based intelligent recognition system for hidden dangers at construction sites. The notice states that algorithm text was incorrectly placed in the original article and provides the correct full Algorithm 1 pseudocode for the YOLO-CGBSE method, including inputs (images and bounding box parameters) and outputs (class probabilities and predicted bounding box coordinates). It also specifies the training and evaluation stages, with GA-optimized YOLO hyperparameters, checkpoint saving, and a publisher apology.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages1

Reading Time3 min

AccessFree to download

Correction

Provided correct Algorithm 1 pseudocode

Training stage and outputs

Reference

OPEN ACCESS Citation: The PLOS One Staff (2026) Correction: YOLO-based intelligent recognition system for hidden dangers at construction sites. PLoS One 21(1): e0340613. [https://doi](https://doi). org/10.1371/journal.pone.0340613 Published: January 6, 2026 Copyright: © 2026 The PLOS One Staff. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. CORRECTION Correction: YOLO-based intelligent recognition system for hidden dangers at construction sites The PLOS One Staff The algorithm text appears incorrectly in the article, as it should have appeared in sequence. The correct full Algorithm are: Algorithm 1 Pseudocode for the YOLO-CGBSE algorithm Input: images Img, Bounding box coordinates Bx, By, width Bw, height Bh Output: Class probabilities Pc and Predicted Bounding box coordinates 1: Initialize: Img = Img_train(80%) + Img_val(20%); 2: batch_size = 24; 3: epochs E = 200; 4: Generation G = 200; 5: for i = 1 : G do 6: for j = 1 : batch_size do 7: Load yolo hyperparameters configuration file; 8: Run Genetic Algorithm (GA) to obtain best hyperparameter values; 9: end for 10: end for 11: Training Stage: 12: for i = 1 : E do 13: Load optimized yolo training parameters from GA; 14: Training on the Img_train with the YOLO-CGBSE; 15: Evaluating algorithm using Img_val; 16: end for 17: save optimal checkpoint bestweight .pt The publisher apologizes for the errors. Reference 1. Li H, Jin P, Zhan L, Yan W, Guo S, Sun S. YOLO-based intelligent recognition system for hidden dangers at construction sites. PLoS One. 2025;20(9):e0332042 . [https://doi.org/10.1371/journal](https://doi.org/10.1371/journal). pone.0332042 PMID: 40961170 PLOS One | [https://doi.org/10.1371/journal.pone.0340613](https://doi.org/10.1371/journal.pone.0340613) January 6, 2026 1 / 1

FAQs

What was corrected in the original PLOS One article?

The algorithm text in the original article appeared incorrectly and should have appeared in sequence. The correction notice provides the correct full Algorithm 1 pseudocode.

What inputs and outputs does Algorithm 1 describe?

Inputs include images and bounding box coordinates/size parameters (Bx, By, Bw, Bh). Outputs include class probabilities (Pc) and predicted bounding box coordinates.

How are YOLO hyperparameters optimized in the corrected algorithm?

A genetic algorithm (GA) is run to obtain the best hyperparameter values, which are then used to train the YOLO model on the training image set and evaluate it on the validation set.

Sep 29, 2026

Upload by kopisore