What is the p(x_m|y_n) in the testing stage?

ZhihengCV / Bayesian-Crowd-Counting

Official Implement of ICCV 2019 oral paper Bayesian Loss for Crowd Count Estimation with Point Supervision

330 stars 84 forks source link

What is the p(x_m|y_n) in the testing stage? #4

Closed CassieMai closed 5 years ago

CassieMai commented 5 years ago

Hi. There are no parameters learned for the Bayesian model. In the training stage, p(x_m|y_n) can be obtained from the ground truth density map. But in the testing stage, how we get the p(x_m|y_n)? Did I miss something?

xingxjtu commented 5 years ago

There is no need to get p(x_m|y_n) in the testing stage since it can be eliminated as shown in Eq.(7).

CassieMai commented 5 years ago

@xingxjtu Thank you for your patience. May I ask one more question? What is the meaning of scale likelihood mentioned in the Conclusion? It would be appreciate if you can give a detailed description.