(Domain Adaptation) Semantic Segmentation
SOTA List
GTA5-2-Cityscapes
SYNTHIA-2-Cityscapes
上表中列出的关于无监督域适应语义分割的文章链接和代码库链接
AdaptSegNet Learning to Adapt Structured Output Space for Semantic Segmentation
github
BDL Bidirectional Learning for Domain Adaptation of Semantic Segmentation
github
AdvEnt Adversarial Entropy Minimization for Do...
DA调研
DA Classification
Transferable Semantic Augmentation for Domain Adaptation
会议: CVPR 2021
作者:北京理工大学
代码:github
Motivation and Contribution
作者认为深度学习网络一定程度上可以使得特征线性化(原文中 deep networks excel at disentangling the underlying
factors of data variation and linearizing the deep features ),某些方向维度上的改变往往和一些语义上的转换相关,比如背景变换,...
计算ICC
Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencoder
这篇文章提到了一个评估特征提取器的泛化能力的指标ICC, 准确来说是Statistical analysis of discriminability。
The ICC is defined as the ratio of inter-class variation and the intraclass variation.
绘制AWA2的ICC(L2正则化特征)
import pickle as pkl
import torch as t
import n...
Feature Transformation Ensemble Model with Batch Spectral Regularization for Cross-Domain Few-Shot Classification
Feature Transformation Ensemble Model with Batch Spectral Regularization for Cross-Domain Few-Shot Classification
作者:Bingyu Liu, Zhen Zhao, Zhenpeng Li, Jianan Jiang, Yuhong Guo, Jieping Ye ; AI Tech, DiDi ChuXing
本文的主要工作
propose a feature transformation ensemble model with batch spectral regularization for the Crossdomain few-...
Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencoder 2021_ICCV
Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencoder
论文的工作
提出了一种泛化能力的模型noise-enhanced supervised autoencoder (NSAE) 用于解决少样本学习中的跨域问题。
NSAE应用并改进了 Supervised autoencoder[1], 并将其用在模型的两阶段训练当中,提高模型的泛化能力。
motivation
our observation is that generalizatio...
89 post articles, 12 pages.