{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/illumination-aware-faster-r-cnn-for-robust","title":"Illumination-aware Faster R-CNN for Robust Multispectral Pedestrian Detection","arxiv_id":"1803.05347","date":"2018-03-14","proceeding":null,"authors":["Chengyang Li","Dan Song","Ruofeng Tong","Min Tang"],"abstract":"Multispectral images of color-thermal pairs have shown more effective than a\nsingle color channel for pedestrian detection, especially under challenging\nillumination conditions. However, there is still a lack of studies on how to\nfuse the two modalities effectively. In this paper, we deeply compare six\ndifferent convolutional network fusion architectures and analyse their\nadaptations, enabling a vanilla architecture to obtain detection performances\ncomparable to the state-of-the-art results. Further, we discover that\npedestrian detection confidences from color or thermal images are correlated\nwith illumination conditions. With this in mind, we propose an\nIllumination-aware Faster R-CNN (IAF RCNN). Specifically, an Illumination-aware\nNetwork is introduced to give an illumination measure of the input image. Then\nwe adaptively merge color and thermal sub-networks via a gate function defined\nover the illumination value. The experimental results on KAIST Multispectral\nPedestrian Benchmark validate the effectiveness of the proposed IAF R-CNN.","url_abs":"http://arxiv.org/abs/1803.05347v2","url_pdf":"http://arxiv.org/pdf/1803.05347v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"multispectral-object-detection","task_name":"Multispectral Object Detection"},{"task_slug":"pedestrian-detection","task_name":"Pedestrian Detection"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"faster-r-cnn","method_name":"Faster R-CNN"},{"method_slug":"rpn","method_name":"RPN"},{"method_slug":"roipool","method_name":"RoIPool"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multispectral-object-detection-on-kaist","task":"Multispectral Object Detection","dataset":"KAIST Multispectral Pedestrian Detection Benchmark","model":"IAFR-CNN","rank_in_archive_order":12,"of":17,"metrics":{"All Miss Rate":"44.23"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1803.05347","atlas_url":"https://app.syntology.ai/?focus=1803.05347","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}