{"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/when-pedestrian-detection-meets-multi-modal","title":"When Pedestrian Detection Meets Multi-Modal Learning: Generalist Model and Benchmark Dataset","arxiv_id":"2407.10125","date":"2024-07-14","proceeding":null,"authors":["Yi Zhang","Wang Zeng","Sheng Jin","Chen Qian","Ping Luo","Wentao Liu"],"abstract":"Recent years have witnessed increasing research attention towards pedestrian detection by taking the advantages of different sensor modalities (e.g. RGB, IR, Depth, LiDAR and Event). However, designing a unified generalist model that can effectively process diverse sensor modalities remains a challenge. This paper introduces MMPedestron, a novel generalist model for multimodal perception. Unlike previous specialist models that only process one or a pair of specific modality inputs, MMPedestron is able to process multiple modal inputs and their dynamic combinations. The proposed approach comprises a unified encoder for modal representation and fusion and a general head for pedestrian detection. We introduce two extra learnable tokens, i.e. MAA and MAF, for adaptive multi-modal feature fusion. In addition, we construct the MMPD dataset, the first large-scale benchmark for multi-modal pedestrian detection. This benchmark incorporates existing public datasets and a newly collected dataset called EventPed, covering a wide range of sensor modalities including RGB, IR, Depth, LiDAR, and Event data. With multi-modal joint training, our model achieves state-of-the-art performance on a wide range of pedestrian detection benchmarks, surpassing leading models tailored for specific sensor modality. For example, it achieves 71.1 AP on COCO-Persons and 72.6 AP on LLVIP. Notably, our model achieves comparable performance to the InternImage-H model on CrowdHuman with 30x smaller parameters. Codes and data are available at https://github.com/BubblyYi/MMPedestron.","url_abs":"https://arxiv.org/abs/2407.10125v1","url_pdf":"https://arxiv.org/pdf/2407.10125v1.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":[{"paper_slug":"when-pedestrian-detection-meets-multi-modal","repo_url":"https://github.com/BubblyYi/MMPedestron","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"3d-object-detection","task_name":"3D Object Detection"},{"task_slug":"multispectral-object-detection","task_name":"Multispectral Object Detection"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"pedestrian-detection","task_name":"Pedestrian Detection"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[{"slug":"mmpd-dataset","name":"MMPD-Dataset","full_name":""}],"methods_introduced":[],"results":[{"leaderboard":"/sota/multispectral-object-detection-on-flir-1","task":"Multispectral Object Detection","dataset":"FLIR","model":"MMPedestron","rank_in_archive_order":1,"of":18,"metrics":{"mAP50":"86.4%"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-crowdhuman-full-body","task":"Object Detection","dataset":"CrowdHuman (full body)","model":"MMPedestron","rank_in_archive_order":2,"of":19,"metrics":{"AP":"97.1","mMR":"30.8"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-eventped","task":"Object Detection","dataset":"EventPed","model":"MMPedestron","rank_in_archive_order":1,"of":6,"metrics":{"AP":"79.0"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-inoutdoor","task":"Object Detection","dataset":"InOutDoor","model":"MMPedestron","rank_in_archive_order":1,"of":6,"metrics":{"AP":"65.7"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-stcrowd","task":"Object Detection","dataset":"STCrowd","model":"MMPedestron","rank_in_archive_order":1,"of":6,"metrics":{"AP":"74.9"},"uses_additional_data":false},{"leaderboard":"/sota/pedestrian-detection-on-llvip","task":"Pedestrian Detection","dataset":"LLVIP","model":"MMPedestron","rank_in_archive_order":1,"of":15,"metrics":{"AP":"0.726"},"uses_additional_data":false},{"leaderboard":"/sota/pedestrian-detection-on-mmpd-dataset","task":"Pedestrian Detection","dataset":"MMPD-Dataset","model":"MMPedestron","rank_in_archive_order":1,"of":1,"metrics":{"box mAP":"79.0"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2407.10125","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}