Skip to main content
. 2022 Feb 24;41(8):1961–1974. doi: 10.1109/TMI.2022.3154599

Fig. 3.

Fig. 3.

An overview of FeatureNet. FeatureNet takes multi-view images as input and extracts multi-view motion feature Inline graphic and multi-view shape feature Inline graphic. Panel (a) describes multi-scale feature fusion. Panel (b) shows the 2D encoder Inline graphic, where Inline graphic refers to SAX, 2CH and 4CH views. Panel (c) describes the combination of multi-view features.