And have you done any experiments with different λ in equation(12)？What kind of effect will it have?

XiaLiPKU / EMANet

The code for Expectation-Maximization Attention Networks for Semantic Segmentation (ICCV'2019 Oral)

https://xialipku.github.io/publication/expectation-maximization-attention-networks-for-semantic-segmentation/

GNU General Public License v3.0

680 stars 130 forks source link

And have you done any experiments with different λ in equation(12)？What kind of effect will it have? #6

Closed Dorispaopao closed 4 years ago

XiaLiPKU commented 5 years ago

No. Setting it as 1 is good enough, so I didn't spend time on it. Moreover, I included it in the paper to state the relationship between EMA and K-means.