[{"methodname": "2019120701", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 62, "methoddesc": "model2"}, {"methodname": "2019120702", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 63, "methoddesc": "faster RCNN"}, {"methodname": "2019120703", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 64, "methoddesc": "Adapted fasterRCNN"}, {"methodname": "adapted faster rcnn", "pubcode": "", "compenv": "GPU tesla p40", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.4, "isanonymous": true, "id": 36, "methoddesc": "Based on faster rcnn, some modifications have been done. The backbone is resnet 50. We only use pedestrian to train. "}, {"methodname": "adapted faster rcnn", "pubcode": "", "compenv": "GPU tesla p40", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.4, "isanonymous": true, "id": 36, "methoddesc": "Based on faster rcnn, some modifications have been done. The backbone is resnet 50. We only use pedestrian to train. "}, {"methodname": "adapted faster rcnn", "pubcode": "", "compenv": "GPU tesla p40", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.4, "isanonymous": true, "id": 37, "methoddesc": "Adapted faster rcnn"}, {"methodname": "AF-detector", "pubcode": "Not yet", "compenv": "1080Ti", "ispublic": false, "pubtitle": "Not yet", "publink": "Not yet", "pubauthors": "Not yet", "pubalias": "No", "pubvenues": "Not yet", "compsubsampling": 4.0, "comptime": 0.5, "isanonymous": true, "id": 88, "methoddesc": "anchor-free detector"}, {"methodname": "AF-detector", "pubcode": "Not yet", "compenv": "1080Ti", "ispublic": false, "pubtitle": "Not yet", "publink": "Not yet", "pubauthors": "Not yet", "pubalias": "No", "pubvenues": "Not yet", "compsubsampling": 1.0, "comptime": 0.5, "isanonymous": true, "id": 89, "methoddesc": "anchor-free detector"}, {"methodname": "AF-detector", "pubcode": "", "compenv": "2080Ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.3, "isanonymous": true, "id": 90, "methoddesc": "Anchor-free detector"}, {"methodname": "anchor free detector", "pubcode": "", "compenv": "2080ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.5, "isanonymous": true, "id": 92, "methoddesc": "anchor free detector"}, {"methodname": "anchor free detector", "pubcode": "", "compenv": "2080ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.5, "isanonymous": true, "id": 94, "methoddesc": "anchor free detector"}, {"methodname": "APD", "pubcode": "", "compenv": "2080ti", "ispublic": true, "pubtitle": "", "publink": "https://arxiv.org/abs/1910.09188", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 93, "methoddesc": "DLA-34"}, {"methodname": "Cascade-DAGN", "pubcode": "-", "compenv": "TITAN X", "ispublic": false, "pubtitle": "-", "publink": "-", "pubauthors": "-", "pubalias": "DSLab", "pubvenues": "-", "compsubsampling": 1.0, "comptime": 0.22, "isanonymous": false, "id": 101, "methoddesc": "cascade R-CNN improved version"}, {"methodname": "cascade-x101", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 54, "methoddesc": "cascade-x101"}, {"methodname": "cascade-x101", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 54, "methoddesc": "cascade-x101"}, {"methodname": "CasDAGN-all", "pubcode": "", "compenv": "TITAN X", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "DSLab", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.22, "isanonymous": false, "id": 106, "methoddesc": "cascade R-CNN "}, {"methodname": "CasDAGN1", "pubcode": "", "compenv": "TITAN", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "DSLab", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.22, "isanonymous": false, "id": 105, "methoddesc": "Cascade R-CNN with attention"}, {"methodname": "DAGN", "pubcode": "", "compenv": "TITAN X", "ispublic": true, "pubtitle": "Occluded Pedestrian Detection Techniques by Deformable Attention-Guided Network (DAGN)", "publink": "https://doi.org/10.3390/app11136025", "pubauthors": "Han Xie, Wenqi Zheng and Hyunchul Shin ", "pubalias": "DSLab", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.21, "isanonymous": false, "id": 107, "methoddesc": "cascade R-CNN with Deformable Attention Guided Module"}, {"methodname": "dfgf", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 26, "methoddesc": "dfgf"}, {"methodname": "Dual shot detector", "pubcode": "https://irtizahasan.com/", "compenv": "PyTorch", "ispublic": false, "pubtitle": "charlie ", "publink": "https://irtizahasan.com/", "pubauthors": "Irtiza Hasan and Li Jinpeng", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 79, "methoddesc": "Based on R-CNN Family"}, {"methodname": "F2DNet", "pubcode": "", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, GTX1080Ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 4949.52, "isanonymous": false, "id": 119, "methoddesc": "A two stage object detector which replaces traditional RPN with CSP based focal detection network and add fast suppression head to suppress false positives."}, {"methodname": "F2DNet", "pubcode": "https://github.com/AbdulHannanKhan/F2DNet", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, GTX1080Ti", "ispublic": true, "pubtitle": "F2DNet: Fast Focal Detection Network for Pedestrian Detection", "publink": "https://www.computer.org/csdl/proceedings-article/icpr/2022/09956732/1IHoOhBF9Li", "pubauthors": "Abdul Hannan Khan, Mohsin Munir, Ludger van Elst, Andreas Dengel", "pubalias": "", "pubvenues": "ICPR 2022", "compsubsampling": null, "comptime": 4949.52, "isanonymous": false, "id": 120, "methoddesc": "A two stage anchor free pedestrian detection model, based on HRNet backbone and feature pyramid network, center and scale head followed by fast suppression head."}, {"methodname": "Faster R-CNN", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.3, "comptime": null, "isanonymous": false, "id": 2, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication."}, {"methodname": "Faster R-CNN", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": false, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 7, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication. Trained on night-time data."}, {"methodname": "Faster R-CNN", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 8, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication. Trained on night-time data only."}, {"methodname": "faster-rcnn", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 55, "methoddesc": "faster-rcnn"}, {"methodname": "faster-rcnn", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 55, "methoddesc": "faster-rcnn"}, {"methodname": "Faster-RCNN", "pubcode": "", "compenv": "1080TI", "ispublic": false, "pubtitle": "Faster-RCNN", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.3, "comptime": 2340.0, "isanonymous": false, "id": 84, "methoddesc": "Attempting to reproduce baseline presented in paper."}, {"methodname": "Faster-RCNN", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.3, "comptime": null, "isanonymous": false, "id": 85, "methoddesc": "Reproducing baseline from paper."}, {"methodname": "FasterRCNN with Multi-Stream RCNN", "pubcode": "", "compenv": "Intel Xeon E5-260, Nvidia TitanX", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.1, "comptime": 0.36, "isanonymous": false, "id": 34, "methoddesc": "We use multi-stream RCNNs to handle the proposals given by an RPN. We use multi-scale training at training time. At test time, only one single scale is used to test. We filter samples that are more than 40% occluded or smaller than 20 pixels in height. We adopt the same setting with Adapted FasterRCNN. We use VGG16 as our based network. We use pedestrian and rider to train."}, {"methodname": "FasterRCNN with Multi-Stream RCNN", "pubcode": "", "compenv": "Intel Xeon E5-260, Nvidia TitanX", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.36, "isanonymous": true, "id": 35, "methoddesc": "We use multi-stream RCNNs to handle the proposals given by an RPN. We use multi-scale training at training time. At test time, only one single scale is used to test. We filter samples that are more than 40% occluded or smaller than 20 pixels in height. We adopt the same setting with Adapted FasterRCNN. We use VGG16 as our based network. We use pedestrian and rider to train."}, {"methodname": "FasterRCNN with Multi-Stream RCNN", "pubcode": "", "compenv": "Intel Xeon E5-260, Nvidia TitanX", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.36, "isanonymous": false, "id": 39, "methoddesc": "We use multi-stream RCNNs to handle the proposals given by an RPN. We filter samples that are more than 80% occluded or smaller than 20 pixels in height. We adopt the same setting with Adapted FasterRCNN. We use VGG16 as our based network. We use pedestrian and rider to train."}, {"methodname": "Fields Detector", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 95, "methoddesc": "Use fields for detection"}, {"methodname": "Fields Detector", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 95, "methoddesc": "Use fields for detection"}, {"methodname": "Fields Detector", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 97, "methoddesc": "Use fields for detection"}, {"methodname": "Fields Detector", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 97, "methoddesc": "Use fields for detection"}, {"methodname": "Irtiza and LiJinpeng (IIAI_VOS)", "pubcode": "will be made soon", "compenv": "", "ispublic": true, "pubtitle": "IIAI_VOS_beta", "publink": "http://irtizahasan.com/", "pubauthors": "Irtiza Hasan and Li Jinpeng", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 58, "methoddesc": "Method is based on Cascade RCNN"}, {"methodname": "Lahore", "pubcode": "http://irtizahasan.com/", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 76, "methoddesc": "Based on Cascade Mask-RCNN"}, {"methodname": "LSFM", "pubcode": "", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, RTX3090", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 3200.0, "isanonymous": true, "id": 121, "methoddesc": "Single stage anchor free efficient pedestrian detector for efficient pedestrian detection."}, {"methodname": "LSFM", "pubcode": "", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, RTX3090", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 2133.0, "isanonymous": true, "id": 122, "methoddesc": "Single stage anchor free efficient pedestrian detector for efficient pedestrian detection."}, {"methodname": "LSFM", "pubcode": "", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, V100", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 1717.0, "isanonymous": true, "id": 124, "methoddesc": "Two-stage anchor free pedestrian detector."}, {"methodname": "LSFM", "pubcode": "", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, V100", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 1717.0, "isanonymous": true, "id": 125, "methoddesc": "Two-staged anchor-free pedestrian detector"}, {"methodname": "LSFM", "pubcode": "https://github.com/AbdulHannanKhan/F2DNet", "compenv": "Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz, RTX3090", "ispublic": true, "pubtitle": "Localized Semantic Feature Mixers for Efficient Pedestrian Detection in Autonomous Driving", "publink": "https://openaccess.thecvf.com/content/CVPR2023/papers/Khan_Localized_Semantic_Feature_Mixers_for_Efficient_Pedestrian_Detection_in_Autonomous_CVPR_2023_paper.pdf", "pubauthors": "Abdul Hannan Khan, Mohammed Shariq Nawaz, Andreas Dengel", "pubalias": "", "pubvenues": "CVPR 2023", "compsubsampling": 1.0, "comptime": 1993.0, "isanonymous": false, "id": 126, "methoddesc": "MLP-Mixers based focal detection network with Super Pixel Pyramid Pooling neck, which uses only linear learnable layer for feature enrichment and filtering instead of computationally costly upscaling operations."}, {"methodname": "LSFM", "pubcode": "", "compenv": "A100", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 667.0, "isanonymous": false, "id": 130, "methoddesc": "MLPMixer based Two-stage anchor-free pedestrian detector"}, {"methodname": "LSFM", "pubcode": "", "compenv": "A100", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": 667.0, "isanonymous": false, "id": 131, "methoddesc": "MLPMixer based Pedestrian Detection Network"}, {"methodname": "MDCN-R50", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "Noah's Ark Lab", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 38, "methoddesc": "ignore objects with area pixels < 20 and group objects."}, {"methodname": "model1", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 61, "methoddesc": "base"}, {"methodname": "model2", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 60, "methoddesc": "new training framework"}, {"methodname": "PCN", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 67, "methoddesc": "PCN method"}, {"methodname": "Pedestrian", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 40, "methoddesc": "Pedestrian"}, {"methodname": "Pedestrian", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 40, "methoddesc": "Pedestrian"}, {"methodname": "Pedestrian2", "pubcode": "will be released as soon as the paper is accepted", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 53, "methoddesc": "Large-scale pedestrian detection."}, {"methodname": "Pedestrian2", "pubcode": "will be released as soon as the paper is accepted", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 53, "methoddesc": "Large-scale pedestrian detection."}, {"methodname": "Pedestron ", "pubcode": "https://github.com/hasanirtiza/Pedestron", "compenv": "Tesla V-100", "ispublic": true, "pubtitle": "Generalizable Pedestrian Detection: The Elephant In The Room. ", "publink": "https://arxiv.org/pdf/2003.08799.pdf", "pubauthors": "Irtiza Hasan, Shengcai Liao, Jinpeng Li , Saad Ullah Akram, and Ling Shao", "pubalias": "IIAI, UAE", "pubvenues": "CVPR 2021", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 81, "methoddesc": "Generalizable Pedestrian Detection: The Elephant In The Room. Accepted at CVPR 2021.\r\nhttps://arxiv.org/pdf/2003.08799.pdf"}, {"methodname": "Pedestron", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 99, "methoddesc": "Based on Pedestron"}, {"methodname": "Pedestron", "pubcode": "", "compenv": "one RTX4090", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": 341724.0, "isanonymous": false, "id": 133, "methoddesc": "The Pedestron repo of Irtiza Hasan is used to train a Cascade R-CNN with MobileNetV2 backbone (smaller than the HRNet used by them) on manipulated training data. The training data is manipulated in a way that ambiguous samples are excluded. In my bachelors thesis, I am investigating the impact of ambiguous data on the training but also on the evaluation and LAMR. I am doing all of the experiments on the validation data but I would like to have a comparable LAMR value on the test dataset. On the non-manipulated validation data, the model achieves the following LAMRs: [0.114, 0.2, 0.499, 0.274]. It was only trained for 50 epochs (original Pedestron is something like 147 epochs) and no external data is used for pretraining."}, {"methodname": "Pedestron (retrained)", "pubcode": "", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 109, "methoddesc": "Train Pedestron on ECP-night"}, {"methodname": "R-FCN (with OHEM)", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.3, "comptime": null, "isanonymous": false, "id": 5, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication."}, {"methodname": "Real-time Pedestrian Detector", "pubcode": "http://irtizahasan.com/", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "http://irtizahasan.com/", "pubauthors": "", "pubalias": "Irtiza and LiJinpeng ", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 71, "methoddesc": "Method is based on Cascade-RCNN"}, {"methodname": "sdfsd", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 29, "methoddesc": "sdfdf"}, {"methodname": "SPNet", "pubcode": "will be released as soon as the paper is accepted", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 46, "methoddesc": "SP-NAS for ECP"}, {"methodname": "SPNet", "pubcode": "will be released as soon as the paper is accepted", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 46, "methoddesc": "SP-NAS for ECP"}, {"methodname": "SPNet", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 82, "methoddesc": "based on cascade"}, {"methodname": "SPNet", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 82, "methoddesc": "based on cascade"}, {"methodname": "SPNet w cascade", "pubcode": "https://github.com/huawei-noah/vega/blob/master/docs/en/algorithms/sp-nas.md", "compenv": "", "ispublic": true, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "Chenhan Jiang, Hang Xu, Wei Zhang, Xiaodan Liang, Zhenguo Li", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 83, "methoddesc": "SP-NAS search for ECP dataset based on cascade head"}, {"methodname": "SPNet w cascade", "pubcode": "https://github.com/huawei-noah/vega/blob/master/docs/en/algorithms/sp-nas.md", "compenv": "", "ispublic": true, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "Chenhan Jiang, Hang Xu, Wei Zhang, Xiaodan Liang, Zhenguo Li", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 83, "methoddesc": "SP-NAS search for ECP dataset based on cascade head"}, {"methodname": "SPNet w FPN", "pubcode": "https://github.com/huawei-noah/vega/blob/master/docs/en/algorithms/sp-nas.md", "compenv": "", "ispublic": true, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 44, "methoddesc": "SP-NAS for ECP based on FPN head, other information can be found in [SP-NAS: Serial-to-Parallel Backbone Search for Object Detection]. No validation data is used during training."}, {"methodname": "SPNet w FPN", "pubcode": "https://github.com/huawei-noah/vega/blob/master/docs/en/algorithms/sp-nas.md", "compenv": "", "ispublic": true, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 44, "methoddesc": "SP-NAS for ECP based on FPN head, other information can be found in [SP-NAS: Serial-to-Parallel Backbone Search for Object Detection]. No validation data is used during training."}, {"methodname": "SPNet w FPN", "pubcode": "https://github.com/huawei-noah/vega", "compenv": "", "ispublic": false, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 66, "methoddesc": "SPNAS search for ECP based on FPN, other information can be found in our paper."}, {"methodname": "SPNet w FPN", "pubcode": "https://github.com/huawei-noah/vega", "compenv": "", "ispublic": false, "pubtitle": "SP-NAS: Serial-to-Parallel Backbone Search for Object Detection", "publink": "http://openaccess.thecvf.com/content_CVPR_2020/papers/Jiang_SP-NAS_Serial-to-Parallel_Backbone_Search_for_Object_Detection_CVPR_2020_paper.pdf", "pubauthors": "", "pubalias": "Huawei Noah AI Theory", "pubvenues": "CVPR2020", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 66, "methoddesc": "SPNAS search for ECP based on FPN, other information can be found in our paper."}, {"methodname": "SS2CM-R-FCN", "pubcode": "None", "compenv": "1080Ti", "ispublic": false, "pubtitle": "No", "publink": "No", "pubauthors": "No", "pubalias": "", "pubvenues": "No", "compsubsampling": 0.78, "comptime": 0.08, "isanonymous": false, "id": 110, "methoddesc": "R-FCN with semantic segmentation as prior"}, {"methodname": "SS2CM-R-FCN", "pubcode": "", "compenv": "1080Ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 0.78, "comptime": 0.08, "isanonymous": false, "id": 111, "methoddesc": "R-FCN with semantic segmentation as prior"}, {"methodname": "SS2CM-R-FCN", "pubcode": "", "compenv": "1080Ti", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 0.78, "comptime": 0.08, "isanonymous": false, "id": 112, "methoddesc": "cascade R-FCN with semantic segmentation as auxliary information"}, {"methodname": "SSD", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 6, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication."}, {"methodname": "SSD", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 74, "methoddesc": "ssd baseline"}, {"methodname": "ssss", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 96, "methoddesc": "fa"}, {"methodname": "ssss", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 96, "methoddesc": "fa"}, {"methodname": "SwapNet", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 42, "methoddesc": "swapnet 1315127 2x mdcn"}, {"methodname": "SwapNet", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 42, "methoddesc": "swapnet 1315127 2x mdcn"}, {"methodname": "SwapNetDB2", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 43, "methoddesc": "1315127 5127 mdcn"}, {"methodname": "SwapNetDB2", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": 1.0, "comptime": null, "isanonymous": true, "id": 43, "methoddesc": "1315127 5127 mdcn"}, {"methodname": "Test", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 100, "methoddesc": "Testing submission"}, {"methodname": "Torchvision Faster-RCNN", "pubcode": "https://github.com/pytorch/vision/blob/master/torchvision/models/detection/faster_rcnn.py", "compenv": "", "ispublic": true, "pubtitle": "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", "publink": "https://papers.nips.cc/paper/5638-faster-r-cnn-towards-real-time-object-detection-with-region-proposal-networks", "pubauthors": "Shaoqing Ren, Kaiming He, Ross Girshick, Jian Sun", "pubalias": "", "pubvenues": "NIPS", "compsubsampling": 1.3, "comptime": null, "isanonymous": false, "id": 87, "methoddesc": "Reproduction of original baseline with correct scaling factor of 1.3 with the PyTorch Torchvision Faster-RCNN model."}, {"methodname": "Two-Shot Detector", "pubcode": "http://irtizahasan.com/", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "Irtiza and LI", "pubalias": "Irtiza and LiJinpeng ", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": true, "id": 75, "methoddesc": "Based on Cascade RCNN"}, {"methodname": "wolverine-test1", "pubcode": "nothing", "compenv": "v100", "ispublic": false, "pubtitle": "nothing", "publink": "nothing", "pubauthors": "nothing", "pubalias": "wolverine", "pubvenues": "nothing", "compsubsampling": 1.0, "comptime": 1.0, "isanonymous": false, "id": 127, "methoddesc": "It is a simple object detection network"}, {"methodname": "wolverine-test2", "pubcode": "nothing", "compenv": "v100", "ispublic": false, "pubtitle": "nothing", "publink": "nothing", "pubauthors": "nothing", "pubalias": "wolverine", "pubvenues": "nothing", "compsubsampling": 1.0, "comptime": 1.0, "isanonymous": false, "id": 128, "methoddesc": "This is a simple object detection model"}, {"methodname": "YOLO-Based", "pubcode": "https://github.com/persolo", "compenv": "Tesla K80", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": 571.975, "isanonymous": true, "id": 113, "methoddesc": "Persolo: A YOLO-Based Model trained on Open Image and MOT20 detection datasets .\r\nCross-validation Experiment on ECP."}, {"methodname": "YOLO-Based", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": 571.975, "isanonymous": true, "id": 117, "methoddesc": "Persolo"}, {"methodname": "YOLOv3", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes", "publink": "https://ieeexplore.ieee.org/document/8634919", "pubauthors": "Braun, Markus and Krebs, Sebastian and Flohr, Fabian B. and Gavrila, Dariu M.", "pubalias": "", "pubvenues": "IEEE Transactions on Pattern Analysis and Machine Intelligence", "compsubsampling": 1.0, "comptime": null, "isanonymous": false, "id": 3, "methoddesc": "One of the baseline methods evaluated in the EuroCity Persons Benchmark publication."}, {"methodname": "YOLOv3", "pubcode": "", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "SMR", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 50, "methoddesc": "based algorithm : yolov3-416\r\n\r\n\r\ndata augmentation : jitter=.3, hue=.1, sat=1.5"}, {"methodname": "yolov3 tiny", "pubcode": "", "compenv": "", "ispublic": false, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 30, "methoddesc": "test"}, {"methodname": "YOLOv3-spp", "pubcode": "", "compenv": "", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "", "pubvenues": "", "compsubsampling": null, "comptime": null, "isanonymous": false, "id": 57, "methoddesc": "used yolov3-spp"}, {"methodname": "YOLOv3_640", "pubcode": "", "compenv": "Nvidia TitanX", "ispublic": true, "pubtitle": "", "publink": "", "pubauthors": "", "pubalias": "HUI_Tsinghua-Daimler Joint Research VRU3-B Project", "pubvenues": "", "compsubsampling": 1.0, "comptime": 0.1, "isanonymous": false, "id": 31, "methoddesc": "One of the test user for the EuroCity Persons Benchmark. Trained on day-time data only. The same training set with the published ECP PAMI2019 paper, height\u2208[20\uff0c\u221e]\uff0c occlusion\u2208[0, 80].\r\nBased on YOLOv3 c++ version (https://github.com/pjreddie/darknet/wiki/YOLO:-Real-Time-Object-Detection\r\n), configurations:  640*640, k-means anchor (k=9), random=1, lr = 1e-03, policy=steps 40000,45000, max_batches=50200.\r\n"}]