As you can see in the picture, I have to determine what the unstandardized coefficients (B) are, but I don't know how and which formula I have to use. I thought about using the standardized coefficients to calculate the unstandardized ones.
Can someone help me ?
The general formula to transition between standardized coefficients, $\beta_1$, and unstandardized coefficients, $b_1$, is: $$\beta_1 = b_1 · \frac{s_1}{s_y}$$ where $s_1$ represents the standard deviation of the first variable in the model (say $x_1$), and $s_y$ represents the standard deviation for the dependent variable for the model.
However, if the only information you have been provided to answer this question is that which you’ve shown here, then this strategy won't work. (You don't have the sample statistics.)
In this case, it appears that you are being asked to utilize another connection in the data provided to answer the question. As this may be a homework question, I will suggest thinking about the connection between the $t$-ratio, the coefficients (the B column), and the corresponding standard error.
Happy to help more if you have a question about this relationship.
