1. Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection paper:https://openaccess.thecvf.com/content/CVPR2021/html/Li_Few-Shot_Object_Detection_via_Classification_Refinement_and_Distractor_Retreatment_CVPR_2021_paper.html code:None 1.1 Motivation 数据稀缺(data scarcity)是FSOD中的难点。
Few-Shot Object Detection with Attention-RPN and Multi-Relation Detectorarxiv.org/abs/1908.01998
few-shot object detection(小样本目标检测)广泛应用于数据有限的条件下,之前很多团队的研究成果聚焦于小样本种类(categories)的表现,旷视研究团队认为在真实应用场景下,测试样本可能包含任何目标物体,因而检测所有类别(classes)至关重要,这需要小样本检测器能够在没有遗忘的条件下学习新的概念(目标)。旷视团队提出了Retent...
CVPR 2021 FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding,程序员大本营,技术文章内容聚合第一站。
Code for CVPR 2022 Oral paper: 'Few-Shot Object Detection with Fully Cross-Transformer' - GuangxingHan/FCT
(Few-shot Detection)Review: Fews-shot Object Detection Feature Reweighting,程序员大本营,技术文章内容聚合第一站。
<p id="abspara0010" view="all"> Traditional object detectors based on deep learning rely on plenty of labeled samples, which are expensive to obtain. Few-shot object detection (FSOD) attempts to solve this problem, learning detection objects from a few l
One-shot learning by inverting a compositional causal process NIPS; (2015) Google Scholar [11] Q. Fan, et al. Few-shot object detection with attention-RPN and multi-relation detector 2020 IEEE/CVF conference on computer vision and pattern recognition (CVPR) IEEE (2020) Google Scholar [12] ...
CVPR2020《Frustratingly simple few-shot object detection. 》提出的TFA方法是基于两阶段的fine-tune指出了小样本目标检测改进方面的巨大潜力。 ECCV2020《Multi-scale positive sample refinement for few-shot object detection》提出的MPSR在TFA的基础上研究了小样本尺度分布与原始样本不同的问题,通过图片金字塔和FPN相...
这是发表在CVPR2022上关于小样本检测的一篇论文,在作者看来,以往小样本检测方法大致可以分为俩类:single-branch方法和two-branch方法;前者通常是基于Faster RCNN进行finetuned,需构建multi-class classifier;但该方法针对shot比较少例如1-shot时,较为容易出现过拟合情况;而后者通常时构建siamese网络,分别同时提取query特...