Liaojiajia-Blog

Unknown-Aware Object Detection: Learning What You Don’t Know from Videos in the Wild1. Motivation2. Overview of the object detection framework2.1 Spatial-Temporal Unknown Distillation2.1.1 Spatial unknown distillation2.1.2 Temporal unknown distillation2.1.

论文链接：https://arxiv.org/abs/2011.09094开源代码：https://github.com/dddzg/up-detrPPT参考链接：https://www.bilibili.com/video/BV1o54y187vG?p=4&t=93

2021 CVPR 2D目标检测1. You Only Look One-level Featurepaper: https://arxiv.org/abs/2103.09460code: https://github.com/megvii-model/YOLOF描述：在单阶段目标检测方面一次突破性的创新，它针对单阶段目标检测中的FPN(特征金字塔)进行了深入的分析并得出：FPN最重要的成分是分而治之的处理思路缓解了优化难问题。针对FPN的多尺度特征、分而治之思想分别提出了Dilated编码器提升特

SyNet1. Motivation2. Method2.1. Object detecion论文链接：开源代码：https://github.com/mertalbaba/SyNet1. Motivation无人机航拍图像与自然图像的拍摄方式不同，导致了以下三个问题：缺乏大目标方差的大型无人机图像数据集；无人机航拍图像有较大的尺度方差和方向变化；地面和空中图像在纹理和形状特征上存在差异。目前已发展的单阶段和多阶段检测器各有优缺点，将每种解决方案的优势结合起来可以达到更强的检测效果

SEAMSelf-supervised Equivariant Attention Mechanism for Weakly Supervised Semantic Segmentation 论文阅读分享PPT

Sparse R-CNNSparse RCNN论文阅读分享PPT

PolarDet1 Introduction2 Approach3 Experiments4 Conclusion论文地址：https://arxiv.org/pdf/2010.08720.pdf1 Introduction2 Approach3 Experiments4 Conclusion

HoughNet1 Introduction2 Approach3 Experiments4 Conclusion论文地址：https://arxiv.org/abs/2007.02355开源代码：https://github.com/nerminsamet/houghnet其他解读：https://mp.weixin.qq.com/s/d-Kz-8KqcB9Afyzva56Ijg本文亮点：1 Introduction2 Approach3 Experiments4 Conclusio

CPNDet1 Introduction论文地址：https://arxiv.org/abs/2007.13816开源代码：https://github.com/Duankaiwen/CPNDet本文亮点：1.1 Introduction

BorderDet 用边界特征做检测Abstract1 Introduction2 Related Work3 The Proposed Approach4 Experiments5 Conclusion论文地址：https://arxiv.org/abs/2007.11056开源代码：https://github.com/Megvii-BaseDetection/BorderDet作者解读：https://zhuanlan.zhihu.com/p/163044323本文亮点：Abstract

@[TOC](论文地址：https://arxiv.org/abs/1912.02252开源代码： https://github.com/KevinKecc/MAL本文亮点：Abstract1 Introduction2 Related Work3 Method3.1 Network Architecture3.2 Feature Selection Module3.3 Dynamic Refinement Head3.4 SKU110K-R Dataset4 Experiment

@[TOC](

DRNAbstract1 Introduction2 Related Work3 Method3.1 Network Architecture3.2 Feature Selection Module3.3 Dynamic Refinement Head3.4 SKU110K-R Dataset4 Experiments4.1 Experimental Results4.2 Ablation Study5 Conclusion论文地址：https://arxiv.org/pdf/2005.09973.pd

SCRDet++PrefaceAbstract1 INTRODUCTION2 RELATED WORK2.1 Horizontal Region Object Detection2.2 Arbitrary-Oriented Object Detection2.3 Image Denoising2.4 Small Object Detection3 THE PROPOSED METHOD3.1 Approach Overview3.2 Instance-level Feature Map Denoising3

EfficientDet论文：https://arxiv.org/abs/1911.09070(非官方)开源代码：(1) Pytorch版：https://github.com/toandaominh1997/EfficientDet.Pytorch(2) Keras&&TensorFlow版：https://github.com/xuannianz/EfficientDet...

EfficientNet 论文：https://arxiv.org/abs/1905.11946EfficientNet 开源代码：https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet第三方实现的PyTorch代码：https://github.com/lukemelas/EfficientNet-...

CentripetalNet前言1、Background Introduction2、Network Architecture3、Experiments4、Conclusions5、References前言论文地址：https://arxiv.org/pdf/2003.09119.pdf开源代码：https://github.com/KiveeDong/CentripetalNet简介：...

目标检测 | Clustered Object Detection in Aerial Images论文阅读一、背景介绍二、网络结构2.1 总体框架2.2 Cluster Region Extraction2.2.1 CPNet2.2.2 ICM2.3 Fine Detection on Cluster Chip2.3.1 ScaleNet2.3.2 PP module2.4 Fusion O...

ICCV2019|论文阅读 SCRDet：Towards More Robust Detection for Small, Cluttered and Rotated Objects前言论文地址：https://arxiv.org/abs/1811.07126开源代码：https://github.com/DetectionTeamUCAS一、背景介绍目标检测计算机视觉的基石。尽管现在...