Enhancing object detection for visually impaired integrating YOLOv8 with spiking EfficientDet.
Saved in:
| Title: | Enhancing object detection for visually impaired integrating YOLOv8 with spiking EfficientDet. |
|---|---|
| Authors: | Biradar, Vinod1,2 (AUTHOR) vinnu151986@gmail.com, Gull, Karuna C.2 (AUTHOR) karunagull74@gmail.com |
| Source: | Multimedia Tools & Applications. Oct2025, Vol. 84 Issue 35, p43721-43749. 29p. |
| Subjects: | People with visual disabilities, Assistive technology, Optimization algorithms, Image segmentation, Object recognition (Computer vision), Computer vision |
| Abstract: | Object detection is a significant method in computer vision, mainly for detecting objects belonging to different classes in an image. Its applications are widely spread to video surveillance, human tracking, autonomous vehicles, and assistive technologies for sight-impaired individuals. However, existing methods suffer from accurately identifying an object and its respective classes due to misidentifying or failing to identify smaller objects. Therefore, this research proposes a new object detection model called You Only Look Once version 8 with Spiking EfficientDet (Yv8SED) that improves detection accuracy, reducing wrong classification with particular emphasis on smaller objects. This model is beneficial in assistive devices for the navigation and safety of a visually-impaired person. The Yv8SED method provides excellent object detection performance with reduced time and cost, making it an efficient methodology for different object detection tasks. The segmentation process is improved through SegNet, which efficiently separates the object from the image. Moreover, the Hippopotamus Optimization Algorithm (HOA) is used to optimize error parameters in the proposed method, thereby improving efficiency and robustness in the object detection process. Experimental results proved the efficiency of the proposed method, with mAP at 95% and mean Average Recall (mAR) at 93% on the COCO dataset and mAP and mAR at 95% and 98%, respectively, on the VOC dataset. These results verify the efficacy and reliability of the proposed method for object detection, particularly for enhancing assistive technologies for the blind and visually impaired. [ABSTRACT FROM AUTHOR] |
| Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Object detection is a significant method in computer vision, mainly for detecting objects belonging to different classes in an image. Its applications are widely spread to video surveillance, human tracking, autonomous vehicles, and assistive technologies for sight-impaired individuals. However, existing methods suffer from accurately identifying an object and its respective classes due to misidentifying or failing to identify smaller objects. Therefore, this research proposes a new object detection model called You Only Look Once version 8 with Spiking EfficientDet (Yv8SED) that improves detection accuracy, reducing wrong classification with particular emphasis on smaller objects. This model is beneficial in assistive devices for the navigation and safety of a visually-impaired person. The Yv8SED method provides excellent object detection performance with reduced time and cost, making it an efficient methodology for different object detection tasks. The segmentation process is improved through SegNet, which efficiently separates the object from the image. Moreover, the Hippopotamus Optimization Algorithm (HOA) is used to optimize error parameters in the proposed method, thereby improving efficiency and robustness in the object detection process. Experimental results proved the efficiency of the proposed method, with mAP at 95% and mean Average Recall (mAR) at 93% on the COCO dataset and mAP and mAR at 95% and 98%, respectively, on the VOC dataset. These results verify the efficacy and reliability of the proposed method for object detection, particularly for enhancing assistive technologies for the blind and visually impaired. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 13807501 |
| DOI: | 10.1007/s11042-025-20872-5 |
