Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{194825,
        author = {Dr. V. Sai Priya and Ch. Sowjanya and G. Venkata Navya and Ch. Amrutha and Ch. Sharvani},
        title = {Multi-object detectionA computer vision approach for efficient object localization and recognition},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {8061-8066},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=194825},
        abstract = {The performance of conventional hand-crafted, feature-based and multi-stage approaches was hampered by problems such as occlusions, size fluctuations, illumination changes, and crowded situations, which make multi-object recognition in computer vision still difficult. Convolutional Neural Networks (CNNs), in particular, made it possible to automatically extract hierarchical features from raw photos with the advent of deep learning, significantly increasing accuracy. The You Only Look Once (YOLO) architecture transformed the field by presenting object identification as a single, cohesive regression issue. This allowed for genuine real-time end-to-end performance by predicting bounding boxes, abjectness scores, and class probabilities in a single forward pass. This study suggests a real-time multi-object detection system that is solely based on an improved YOLO framework with a CNN backbone for reliable multi-scale feature extraction. It achieves high localization and classification accuracy even in intricate multi-object situations while preserving low-latency inference appropriate for robotics, autonomous driving, and surveillance applications. According to experimental results, our simplified CNN + YOLO method successfully strikes a compromise between speed, accuracy, and recall without the need for further post-processing stages.},
        keywords = {Deep learning, Computer Vision, object detection, Yolo and CNN, Hand-Crafted, Feature-Based and Multi-Stage, fluctuations, illumination.},
        month = {March},
        }