{"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/adaptive-nms-refining-pedestrian-detection-in","title":"Adaptive NMS: Refining Pedestrian Detection in a Crowd","arxiv_id":"1904.03629","date":"2019-04-07","proceeding":"CVPR 2019 6","authors":["Songtao Liu","Di Huang","Yunhong Wang"],"abstract":"Pedestrian detection in a crowd is a very challenging issue. This paper\naddresses this problem by a novel Non-Maximum Suppression (NMS) algorithm to\nbetter refine the bounding boxes given by detectors. The contributions are\nthreefold: (1) we propose adaptive-NMS, which applies a dynamic suppression\nthreshold to an instance, according to the target density; (2) we design an\nefficient subnetwork to learn density scores, which can be conveniently\nembedded into both the single-stage and two-stage detectors; and (3) we achieve\nstate of the art results on the CityPersons and CrowdHuman benchmarks.","url_abs":"http://arxiv.org/abs/1904.03629v1","url_pdf":"http://arxiv.org/pdf/1904.03629v1.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":"object-detection","task_name":"Object Detection"},{"task_slug":"pedestrian-detection","task_name":"Pedestrian Detection"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"adaptive-nms","method_name":"Adaptive NMS"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"fpn","method_name":"FPN"},{"method_slug":"faster-r-cnn","method_name":"Faster R-CNN"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"rpn","method_name":"RPN"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"roipool","method_name":"RoIPool"},{"method_slug":"sgd-with-momentum","method_name":"SGD with Momentum"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"step-decay","method_name":"Step Decay"},{"method_slug":"weight-decay","method_name":"Weight Decay"}],"datasets_introduced":[],"methods_introduced":[{"slug":"adaptive-nms","name":"Adaptive NMS","full_name":"Adaptive NMS"}],"results":[{"leaderboard":"/sota/object-detection-on-crowdhuman-full-body","task":"Object Detection","dataset":"CrowdHuman (full body)","model":"Adaptive NMS (Faster RCNN, ResNet50)","rank_in_archive_order":18,"of":19,"metrics":{"AP":"84.71","mMR":"49.73"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1904.03629","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}