2016/03/01 by JT Turner, Kalyan Moy Gupta, Turner, JT +6
Computer Science · Engineering · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Artificial intelligence #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Convolutional neural network #Data mining #FOS: Computer and information sciences #Industrial Vision Systems and Defect Detection #Machine learning #Object (grammar) #Object detection #Pattern recognition (psychology) #cs.CV
paper · pdf · doi:10.48550/arxiv.1603.00502
9 pages, 5 figures, 3 tables
arxiv created 2016/03/01 · openalex publication_date 2016/03/01 · arxiv updated 2016/03/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Although recent advances in regional Convolutional Neural Networks (CNNs)\nenable them to outperform conventional techniques on standard object detection\nand classification tasks, their response time is still slow for real-time\nperformance. To address this issue, we propose a method for region proposal as\nan alternative to selective search, which is used in current state-of-the art\nobject detection algorithms. We evaluate our Keypoint Density-based Region\nProposal (KDRP) approach and show that it speeds up detection and\nclassification on fine-grained tasks by 100% versus the existing selective\nsearch region proposal technique without compromising classification accuracy.\nKDRP makes the application of CNNs to real-time detection and classification\nfeasible.\n