2019/08/19 by Federico Magliani, Magliani, Federico, Laura Sani +5
Computer Science · #Advanced Image and Video Retrieval Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Image Retrieval and Classification Techniques #Recommender Systems and Techniques
paper · pdf · doi:10.48550/arxiv.1908.06896
openalex publication_date 2019/08/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Several computer vision and artificial intelligence projects are nowadays\nexploiting the manifold data distribution using, e.g., the diffusion process.\nThis approach has produced dramatic improvements on the final performance\nthanks to the application of such algorithms to the kNN graph. Unfortunately,\nthis recent technique needs a manual configuration of several parameters, thus\nit is not straightforward to find the best configuration for each dataset.\nMoreover, the brute-force approach is computationally very demanding when used\nto optimally set the parameters of the diffusion approach. We propose to use\ngenetic algorithms to find the optimal setting of all the diffusion parameters\nwith respect to retrieval performance for each different dataset. Our approach\nis faster than others used as references (brute-force, random-search and PSO).\nA comparison with these methods has been made on three public image datasets:\nOxford5k, Paris6k and Oxford105k.\n