motivation:传统的few-shot分类将每一类表示为一个特征点,在样本较小的情况下具有较大的噪声,能否把每一类的原型表示为一个分布? 方法:将每一个点表示为一个分布,约束分布之间的分类,以及同一类的样本的分布类似。 Collect and Select: Semantic Alignment Metric Learning for Few-Shot Learning motivation:小样本任...
Few-Shot Object Detection via Variational Feature Aggregation (AAAI2023) Jiaming Han,Yuqiang Ren,Jian Ding,Ke Yan,Gui-Song Xia. arXiv preprint. Our code is based onmmfewshot. Setup Installation Here is a from-scratch setup script. conda create -n vfa python=3.8 -y conda activate vfa conda...
Zero-shot Slot Filling with Slot-Prefix Prompting and Attention Relationship DescriptorLuo Qiaoyang; Liu LingqiaoDoodle to Object: Practical Zero-Shot Sketch-Based 3D Shape RetrievalWang Bingrui; Zhou YuanDUET: Cross-modal Semantic Grounding for Contrastive Zero-shot LearningChen Zhuo; Huang Yufeng; ...
One-shot learning of scene locations via feature trajectory transfer, in CVPR, 2016. R. Kwitt, ...
Breaking Immutable: Information-Coupled Prototype Elaboration for Few-Shot Object DetectionAAAI 2023PASCAL VOC & MS COCOPDFCODE Few-Shot Object Detection via Variational Feature AggregationAAAI 2023PASCAL VOC & MS COCOPDFCODE Disentangle and Remerge: Interventional Knowledge Distillation for Few-Shot Object...
Variational Prototyping-Encoder: One-Shot Learning with Prototypical Images motivation:针对路标的小样本分类问题。对真实场景的路标进行分类,可以将图像化的路标图像作为原型?通过将真实场景的路标特征与图像化的路标特征比较,可以得到真实场景的路标特征。在已知原型的基础上,如何拉近query样本和原型特征的距离?
基于语义的方法:《Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders 基于对齐变分自编码器的广义零炮和少图像学习》简介 本文作者训练了两个变分自编码器(VAE),一个用于视觉特征,另一个则用于语义特征。其目的是能够根据潜在的视觉特征重建语义特征,反之亦然。作者表明,使得两个潜在空间...
Few-shot object detection via data augmentation and distribution calibration General object detection has been widely developed and studied over the past few years, while few-shot object detection is still in the exploratory stage. ... S Zhu,K Zhang - 《Machine Vision & Applications》 被引量: ...
A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities 上周读了这篇22年的新综述,感慨论文真的看不完,而且自己看论文也容易过拟合到某个小角落出不来。这篇主要讨论了meta-learning, transfer learning, data augmentation, multimodal相关的few-shot learning方法,读起...
metric learning via variational inference Dense classification and implanting for few-shot learning. CVPR 2019 metric-based [by CDFS] Subspace Networks for Few-shot Classification. arXiv'1905.13613 follow "A Closer Look at Few-shot Classification" 的設定 根據embedded query point 到每個 class subspace...