{"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/scale-aware-trident-networks-for-object","title":"Scale-Aware Trident Networks for Object Detection","arxiv_id":"1901.01892","date":"2019-01-07","proceeding":"ICCV 2019 10","authors":["Yanghao Li","Yuntao Chen","Naiyan Wang","Zhao-Xiang Zhang"],"abstract":"Scale variation is one of the key challenges in object detection. In this work, we first present a controlled experiment to investigate the effect of receptive fields for scale variation in object detection. Based on the findings from the exploration experiments, we propose a novel Trident Network (TridentNet) aiming to generate scale-specific feature maps with a uniform representational power. We construct a parallel multi-branch architecture in which each branch shares the same transformation parameters but with different receptive fields. Then, we adopt a scale-aware training scheme to specialize each branch by sampling object instances of proper scales for training. As a bonus, a fast approximation version of TridentNet could achieve significant improvements without any additional parameters and computational cost compared with the vanilla detector. On the COCO dataset, our TridentNet with ResNet-101 backbone achieves state-of-the-art single-model results of 48.4 mAP. Codes are available at https://git.io/fj5vR.","url_abs":"https://arxiv.org/abs/1901.01892v2","url_pdf":"https://arxiv.org/pdf/1901.01892v2.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":"scale-aware-trident-networks-for-object","repo_url":"https://github.com/chengzhengxin/groupsoftmax-simpledet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"scale-aware-trident-networks-for-object","repo_url":"https://github.com/facebookresearch/detectron2/tree/master/projects/TridentNet/","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"scale-aware-trident-networks-for-object","repo_url":"https://github.com/tusimple/simpledet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"scale-aware-trident-networks-for-object","repo_url":"https://github.com/open-mmlab/mmdetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bottleneck-residual-block","method_name":"Bottleneck Residual Block"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"deformable-convolution","method_name":"Deformable Convolution"},{"method_slug":"dilated-convolution","method_name":"Dilated Convolution"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"randomhorizontalflip","method_name":"Random Horizontal Flip"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"soft-nms","method_name":"Soft-NMS"},{"method_slug":"step-decay","method_name":"Step Decay"},{"method_slug":"tridentnet","method_name":"TridentNet"},{"method_slug":"tridentnet-block","method_name":"TridentNet Block"}],"datasets_introduced":[],"methods_introduced":[{"slug":"tridentnet","name":"TridentNet","full_name":"TridentNet"},{"slug":"tridentnet-block","name":"TridentNet Block","full_name":"TridentNet Block"}],"results":[{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"TridentNet (ResNet-101)","rank_in_archive_order":158,"of":220,"metrics":{"AP50":"63.5","AP75":"45.5","APL":"56.9","APM":"47","APS":"24.9","box AP":"42"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"TridentNet (ResNet-101-Deformable, Image Pyramid)","rank_in_archive_order":105,"of":225,"metrics":{"AP50":"69.7","AP75":"53.5","APL":"60.3","APM":"51.3","APS":"31.8","box mAP":"48.4"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco","task":"Object Detection","dataset":"COCO test-dev","model":"TridentNet (ResNet-101)","rank_in_archive_order":170,"of":225,"metrics":{"AP50":"63.6","AP75":"46.5","APL":"56.6","APM":"46.6","APS":"23.9","box mAP":"42.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1901.01892","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}