UCNet
2000년 도입 · 논문 1편에서 사용
UCNet is a probabilistic framework for RGB-D Saliency Detection that employs uncertainty by learning from the data labelling process. It utilizes conditional variational autoencoders to model human annotation uncertainty and generate multiple saliency maps for each input image by sampling in the latent space.
출처: UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders
소개 논문: UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders
RGB-D Saliency Detection Models · Computer Vision