Object Detection(目标检测神文)(二)

本文汇总了CVPR 2019等会议上关于目标检测的最新研究,包括无锚框方法、YOLO系列改进、以及行人检测的挑战与解决方案。文章探讨了Generalized Intersection over Union作为新的损失函数和评估指标,以及一系列创新的检测框架,如CenterNet、FoveaBox、RepPoints等。此外,还介绍了针对拥挤场景中行人检测的Repulsion Loss和Occlusion-aware R-CNN等技术。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

文章目录

[CVPR2019] Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression

anchor-free

无锚框最近的热点,有机会研究下。

[CVPR2019] Region Proposal by Guided Anchoring

[CVPR2019] Feature Selective Anchor-Free Module for Single-Shot Object Detection

[CVPR2019]CenterNet: Keypoint Triplets for Object Detection

[CVPR2019]Objects as Points

[CVPR2019]CornerNet-Lite: Efficient Keypoint Based Object Detection

[CVPR2019]FoveaBox: Beyond Anchor-based Object Detector

[2019]DuBox: No-Prior Box Objection Detection via Residual Dual Scale Detectors

YOLO

[2019]Spiking-YOLO: Spiking Neural Network for Real-time Object Detection

[CVPR2019]Gaussian YOLOv3: An Accurate and Fast Object Detector Using Localization Uncertainty for Autonomous Driving

[AAAI2019]Gradient Harmonized Single-stage Detector

[2019]Augmentation for small object detection

[2019]SimpleDet: A Simple and Versatile Distributed Framework for Object Detection and Instance Recognition

[2019]BayesOD: A Bayesian Approach for Uncertainty Estimation in Deep Object Detectors

[2019]DetNAS: Neural Architecture Search on Object Detection

  • intro: Chinese Academy of Sciences & Megvii Inc
  • arxiv:
评论 12
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值