UCNet

Introduced by Zhang et al. in UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders

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.

Source: UC-Net: Uncertainty Inspired RGB-D Saliency Detection via Conditional Variational Autoencoders

Read Paper See Code

Papers

Paper	Code	Results	Date	Stars

Tasks

Task	Papers	Share
RGB-D Salient Object Detection	1	33.33%
Saliency Detection	1	33.33%
Thermal Image Segmentation	1	33.33%

Usage Over Time

This feature is experimental; we are continuously improving our matching algorithm.

Components

Component	Type	Add Remove
🤖 No Components Found	You can add them if they exist; e.g. Mask R-CNN uses RoIAlign

Categories

Add Remove

RGB-D Saliency Detection Models