{"url":"/method/fpn","slug":"fpn","name":"FPN","full_name":"Feature Pyramid Network","full_name_withheld":false,"description_markdown":"A **Feature Pyramid Network**, or **FPN**, is a feature extractor that takes a single-scale image of an arbitrary size as input, and outputs proportionally sized feature maps at multiple levels, in a fully convolutional fashion. This process is independent of the backbone convolutional architectures. It therefore acts as a generic solution for building feature pyramids inside deep convolutional networks to be used in tasks like object detection.\r\n\r\nThe construction of the pyramid involves a bottom-up pathway and a top-down pathway.\r\n\r\nThe bottom-up pathway is the feedforward computation of the backbone ConvNet, which computes a feature hierarchy consisting of feature maps at several scales with a scaling step of 2. For the feature\r\npyramid, one pyramid level is defined for each stage. The output of the last layer of each stage is used as a reference set of feature maps. For [ResNets](https://paperswithcode.com/method/resnet) we use the feature activations output by each stage’s last [residual block](https://paperswithcode.com/method/residual-block). \r\n\r\nThe top-down pathway hallucinates higher resolution features by upsampling spatially coarser, but semantically stronger, feature maps from higher pyramid levels. These features are then enhanced with features from the bottom-up pathway via lateral connections. Each lateral connection merges feature maps of the same spatial size from the bottom-up pathway and the top-down pathway. The bottom-up feature map is of lower-level semantics, but its activations are more accurately localized as it was subsampled fewer times.","description_state":"present","introduced_year":null,"introduced_by":{"title":null,"paper":null,"first_author":null,"n_authors":0,"url_abs":null,"archive_paper_url":null},"source":{"url":"http://arxiv.org/abs/1612.03144v2","title":"Feature Pyramid Networks for Object Detection","url_on_a_paper_host":true},"code_snippet_url":"https://github.com/facebookresearch/Detectron/blob/8170b25b425967f8f1c7d715bea3c5b8d9536cd8/detectron/modeling/FPN.py#L117","code_snippet_url_on_a_code_host":true,"categories":[{"area":"Computer Vision","area_id":"computer-vision","collection":"Feature Extractors","url":"/methods/category/feature-extractors","pwc_aliases":[]}],"n_papers_tagged":583,"archive_num_papers":null,"papers_newest_first":[{"paper":"/paper/kg-htc-integrating-knowledge-graphs-into-llms","title":"KG-HTC: Integrating Knowledge Graphs into LLMs for Effective Zero-shot Hierarchical Text Classification","date":"2025-05-08","arxiv_id":"2505.05583","n_code_links":1,"syntology":null},{"paper":null,"title":"PaniCar: Securing the Perception of Advanced Driving Assistance Systems Against Emergency Vehicle Lighting","date":"2025-05-08","arxiv_id":"2505.05183","n_code_links":0,"syntology":null},{"paper":null,"title":"Floating Car Observers in Intelligent Transportation Systems: Detection Modeling and Temporal Insights","date":"2025-04-29","arxiv_id":"2505.02845","n_code_links":0,"syntology":null},{"paper":null,"title":"Class Imbalance Correction for Improved Universal Lesion Detection and Tagging in CT","date":"2025-04-08","arxiv_id":"2504.05591","n_code_links":0,"syntology":null},{"paper":null,"title":"Resting State Functional Connectivity Patterns Associate with Alcohol Use Disorder Characteristics: Insights from the Triple Network Model","date":"2025-04-08","arxiv_id":"2504.06199","n_code_links":0,"syntology":null},{"paper":null,"title":"BBoxCut: A Targeted Data Augmentation Technique for Enhancing Wheat Head Detection Under Occlusions","date":"2025-03-31","arxiv_id":"2503.24032","n_code_links":0,"syntology":null},{"paper":"/paper/event-based-crossing-dataset-ebcd","title":"Event-Based Crossing Dataset (EBCD)","date":"2025-03-21","arxiv_id":"2503.17499","n_code_links":1,"syntology":null},{"paper":null,"title":"YOLO-LLTS: Real-Time Low-Light Traffic Sign Detection via Prior-Guided Enhancement and Multi-Branch Feature Interaction","date":"2025-03-18","arxiv_id":"2503.13883","n_code_links":0,"syntology":null},{"paper":null,"title":"Securing Virtual Reality Experiences: Unveiling and Tackling Cybersickness Attacks with Explainable AI","date":"2025-03-17","arxiv_id":"2503.13419","n_code_links":0,"syntology":null},{"paper":"/paper/walnutdata-a-uav-remote-sensing-dataset-of","title":"WalnutData: A UAV Remote Sensing Dataset of Green Walnuts and Model Evaluation","date":"2025-02-27","arxiv_id":"2502.20092","n_code_links":1,"syntology":null},{"paper":null,"title":"Hybrid Answer Set Programming: Foundations and Applications","date":"2025-02-13","arxiv_id":"2502.09235","n_code_links":0,"syntology":null},{"paper":null,"title":"Fast-COS: A Fast One-Stage Object Detector Based on Reparameterized Attention Vision Transformer for Autonomous Driving","date":"2025-02-11","arxiv_id":"2502.07417","n_code_links":0,"syntology":null},{"paper":"/paper/mhaf-yolo-multi-branch-heterogeneous","title":"MHAF-YOLO: Multi-Branch Heterogeneous Auxiliary Fusion YOLO for accurate object detection","date":"2025-02-07","arxiv_id":"2502.04656","n_code_links":1,"syntology":null},{"paper":null,"title":"Vision-Integrated LLMs for Autonomous Driving Assistance : Human Performance Comparison and Trust Evaluation","date":"2025-02-06","arxiv_id":"2502.06843","n_code_links":0,"syntology":null},{"paper":null,"title":"YOLOv4: A Breakthrough in Real-Time Object Detection","date":"2025-02-06","arxiv_id":"2502.04161","n_code_links":0,"syntology":null},{"paper":null,"title":"SPFFNet: Strip Perception and Feature Fusion Spatial Pyramid Pooling for Fabric Defect Detection","date":"2025-02-03","arxiv_id":"2502.01445","n_code_links":0,"syntology":null},{"paper":null,"title":"Empirical modeling and hybrid machine learning framework for nucleate pool boiling on microchannel structured surfaces","date":"2025-01-28","arxiv_id":"2501.16867","n_code_links":0,"syntology":null},{"paper":null,"title":"Efficient Object Detection of Marine Debris using Pruned YOLO Model","date":"2025-01-27","arxiv_id":"2501.16571","n_code_links":0,"syntology":null},{"paper":"/paper/a-transformer-based-autoregressive-decoder","title":"A Transformer-based Autoregressive Decoder Architecture for Hierarchical Text Classification","date":"2025-01-23","arxiv_id":"2501.13598","n_code_links":1,"syntology":null},{"paper":null,"title":"Dual Scale-aware Adaptive Masked Knowledge Distillation for Object Detection","date":"2025-01-13","arxiv_id":"2501.07101","n_code_links":0,"syntology":null},{"paper":"/paper/cpdr-towards-highly-efficient-salient-object","title":"CPDR: Towards Highly-Efficient Salient Object Detection via Crossed Post-decoder Refinement","date":"2025-01-11","arxiv_id":"2501.06441","n_code_links":0,"syntology":null},{"paper":"/paper/detection-of-body-packs-in-abdominal-ct-scans","title":"Detection of Body Packs in Abdominal CT scans Through Artificial Intelligence","date":"2024-12-26","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/distortion-aware-adversarial-attacks-on","title":"Distortion-Aware Adversarial Attacks on Bounding Boxes of Object Detectors","date":"2024-12-25","arxiv_id":"2412.18815","n_code_links":1,"syntology":null},{"paper":"/paper/comprehensive-multi-modal-prototypes-are","title":"Comprehensive Multi-Modal Prototypes are Simple and Effective Classifiers for Vast-Vocabulary Object Detection","date":"2024-12-23","arxiv_id":"2412.17800","n_code_links":1,"syntology":null},{"paper":"/paper/qtseg-a-query-token-based-architecture-for","title":"QTSeg: A Query Token-Based Architecture for Efficient 2D Medical Image Segmentation","date":"2024-12-23","arxiv_id":"2412.17241","n_code_links":1,"syntology":null},{"paper":null,"title":"Object Detection Approaches to Identifying Hand Images with High Forensic Values","date":"2024-12-21","arxiv_id":"2412.16431","n_code_links":0,"syntology":null},{"paper":null,"title":"Exploring Machine Learning Engineering for Object Detection and Tracking by Unmanned Aerial Vehicle (UAV)","date":"2024-12-19","arxiv_id":"2412.15347","n_code_links":0,"syntology":null},{"paper":null,"title":"HS-FPN: High Frequency and Spatial Perception FPN for Tiny Object Detection","date":"2024-12-13","arxiv_id":"2412.10116","n_code_links":0,"syntology":null},{"paper":"/paper/emov2-pushing-5m-vision-model-frontier","title":"EMOv2: Pushing 5M Vision Model Frontier","date":"2024-12-09","arxiv_id":"2412.06674","n_code_links":1,"syntology":null},{"paper":"/paper/psych-occlusion-using-visual-psychophysics","title":"Psych-Occlusion: Using Visual Psychophysics for Aerial Detection of Occluded Persons during Search and Rescue","date":"2024-12-07","arxiv_id":"2412.05553","n_code_links":1,"syntology":null}],"papers_shown":30,"tasks":[{"task":"/task/object-detection","name":"Object Detection","papers":342},{"task":"/task/object-detection-1","name":"object-detection","papers":311},{"task":"/task/object","name":"Object","papers":192},{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":96},{"task":"/task/instance-segmentation","name":"Instance Segmentation","papers":62},{"task":"/task/segmentation","name":"Segmentation","papers":60},{"task":"/task/image-classification","name":"Image Classification","papers":43},{"task":"/task/image-classification","name":"image-classification","papers":27},{"task":"/task/data-augmentation","name":"Data Augmentation","papers":25},{"task":null,"name":"GPU","papers":25},{"task":"/task/classification-1","name":"Classification","papers":23},{"task":"/task/decoder","name":"Decoder","papers":23},{"task":"/task/knowledge-distillation","name":"Knowledge Distillation","papers":23},{"task":"/task/autonomous-driving","name":"Autonomous Driving","papers":22},{"task":"/task/classification","name":"General Classification","papers":22},{"task":"/task/real-time-object-detection","name":"Real-Time Object Detection","papers":21},{"task":"/task/text-classification","name":"Text Classification","papers":19},{"task":"/task/deep-learning","name":"Deep Learning","papers":18},{"task":"/task/pedestrian-detection","name":"Pedestrian Detection","papers":18},{"task":"/task/text-classification-1","name":"text-classification","papers":18}],"tasks_shown":20,"n_tasks":351,"usage_by_year":[{"year":"2016","papers":1},{"year":"2017","papers":5},{"year":"2018","papers":20},{"year":"2019","papers":82},{"year":"2020","papers":102},{"year":"2021","papers":109},{"year":"2022","papers":99},{"year":"2023","papers":80},{"year":"2024","papers":64},{"year":"2025","papers":21}],"row_source":"embedded","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/fpn"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}