An Innovative Walrus Optimization-Based Extended Yolov9 Model for Object Detection

The problem of object detection (OD) is widely used in visual recognition. It involves detecting and recognizing objects within images and determining their positions. As opposed to humans, for whom it is very easy to identify and distinguish objects, computers encounter difficulties in dealing with factors such as variation in the appearance of the object, overlap between objects, illumination, and cluttered scenes. However, due to advances in deep learning (DL), OD has become highly advanced in enabling automatic methods to achieve a higher degree of efficiency and accuracy. To tackle the above-mentioned challenges, this paper aims to enhance the performance of YOLOv9 through better feature extraction, multi-scale architecture, and preprocessing, while the Walrus optimization algorithm (WaO) will be used as an auxiliary technique. Primarily, the raw images gathered from public sources are pre-processed using the Gabor anisotropic diffusion (GAD) filtering technique, which eliminates noise and improves the image quality. After preprocessing, OD is performed using the Extended You Only Look Once version 9 (Ex-YOLOv9) technique, which recognizes different objects present in the images. Moreover, the parameters of the developed model are tuned using the WaO approach that minimizes the complexities and enhances the model performance. The proposed method is simulated via the Python platform, and various performance measures are validated and compared with conventional frameworks. The WaO-Ex-YOLOv9 framework achieved an overall accuracy of 97.82% across the five evaluated classes, along with a mAP of 95.40%, [email protected]:0.95 of 94.60%, and an IoU of 96.21%.

Authors

Institutions

Publication Details

Journal
International Journal of Software Engineering and Knowledge Engineering
Published
2026-11-18
DOI
https://doi.org/10.1142/s0218194026500956
Primary Topic
Advanced Neural Network Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

An Innovative Walrus Optimization-Based Extended Yolov9 Model for Object Detection

G C R Kartheek, Manoj Challa
1 links
International Journal of Software Engineering and Knowledge Engineering
Advanced Neural Network Applications
article

An Innovative Walrus Optimization-Based Extended Yolov9 Model for Object Detection

G C R Kartheek, Manoj Challa
article en

Abstract

The problem of object detection (OD) is widely used in visual recognition. It involves detecting and recognizing objects within images and determining their positions. As opposed to humans, for whom it is very easy to identify and distinguish objects, computers encounter difficulties in dealing with factors such as variation in the appearance of the object, overlap between objects, illumination, and cluttered scenes. However, due to advances in deep learning (DL), OD has become highly advanced in enabling automatic methods to achieve a higher degree of efficiency and accuracy. To tackle the above-mentioned challenges, this paper aims to enhance the performance of YOLOv9 through better feature extraction, multi-scale architecture, and preprocessing, while the Walrus optimization algorithm (WaO) will be used as an auxiliary technique. Primarily, the raw images gathered from public sources are pre-processed using the Gabor anisotropic diffusion (GAD) filtering technique, which eliminates noise and improves the image quality. After preprocessing, OD is performed using the Extended You Only Look Once version 9 (Ex-YOLOv9) technique, which recognizes different objects present in the images. Moreover, the parameters of the developed model are tuned using the WaO approach that minimizes the complexities and enhances the model performance. The proposed method is simulated via the Python platform, and various performance measures are validated and compared with conventional frameworks. The WaO-Ex-YOLOv9 framework achieved an overall accuracy of 97.82% across the five evaluated classes, along with a mAP of 95.40%, [email protected]:0.95 of 94.60%, and an IoU of 96.21%.

International Journal of Software Engineering and Knowledge Engineering
Visvesvaraya Technological University (IN)
Advanced Neural Network Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon