当前基于深度学习的域适应,主要研究领域以及代表性的文章:
1.分类
Duplex Generative Adversarial Network for Unsupervised Domain Adaptation
Maximum Classifier Discrepancy for Unsupervised Domain Adaptation
2. 安防的行人重识别(Person Re-Identification)
利用计算机视觉技术判断图像或者视频序列中是否存在特定行人的技术。广泛被认为是一个图像检索的子问题。给定一个监控行人图像,检索跨设备下的该行人图像。旨在弥补目前固定的摄像头的视觉局限,并可与行人检测/行人跟踪技术相结合,可广泛应用于智能视频监控、智能安保等领域。
Person Transfer GAN to Bridge Domain Gap for Person Re-Identification
3.自动驾驶背景下的语义分割( Semantic Segmentation)
Conditional Generative Adversarial Network for Structured Domain Adaptation
4.图像深度估计(Depth Estimation)
Single-Image Depth Estimation Based on Fourier Domain Analysis
Cross-Domain Self-Supervised Multi-Task Feature Learning Using Synthetic Imagery
AdaDepth: Unsupervised Content Congruent Adaptation for Depth Estimation
5. 目标检测
Domain Adaptive Faster R-CNN for Object Detection in the Wild