Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated: 1 00:00:00,510 --> 00:00:06,180 Hello and welcome to this Python tutorial we have just defined in the previous Statoil the architecture 2 00:00:06,480 --> 00:00:08,470 of the neural network of the generator. 3 00:00:08,670 --> 00:00:15,600 And now we need to add a tool to our class G which will be the forward function that will forward propagate 4 00:00:15,720 --> 00:00:19,530 the signal inside the whole neural network. 5 00:00:19,530 --> 00:00:23,430 The good news is that after the storm we had in the British Statoil. 6 00:00:23,550 --> 00:00:29,760 Now things are going to get really easy because we will simply use the main object which remains an 7 00:00:29,760 --> 00:00:35,940 object of the sequential class that contains all the different modules to forward propagate the signal 8 00:00:36,020 --> 00:00:37,320 in just one line of code. 9 00:00:37,500 --> 00:00:38,520 Let's do it. 10 00:00:38,520 --> 00:00:40,740 We are going to define a new function. 11 00:00:40,740 --> 00:00:45,180 There are forms starting with death and then we need to specify the name of this function which will 12 00:00:45,180 --> 00:00:51,650 be forward for the forward propagation and this function is going to take two arguments. 13 00:00:51,780 --> 00:00:57,710 Self our object which represents nothing else than the neural network itself. 14 00:00:57,780 --> 00:00:59,720 The neural network of the generator. 15 00:00:59,910 --> 00:01:06,090 And also because as I've just said we're going to use the main object to propagate the signal but since 16 00:01:06,330 --> 00:01:10,060 this main object is a variable attached to our object. 17 00:01:10,230 --> 00:01:15,770 Well we need to input ourselves object here to be able to use self-determine. 18 00:01:15,840 --> 00:01:16,240 All right. 19 00:01:16,260 --> 00:01:23,160 And then the second argument you can see the year is actually going to be the input that is nothing 20 00:01:23,160 --> 00:01:27,160 else than the input of the neural network of the generator. 21 00:01:27,180 --> 00:01:32,520 I remind this input is going to be some random vector of size 100. 22 00:01:32,520 --> 00:01:39,160 That's why we specified 100 here for the size of the input of our first inverse convolution. 23 00:01:39,180 --> 00:01:45,480 This is going to be a random vector input of the generator which will just represent some noise to generate 24 00:01:45,560 --> 00:01:49,680 a fake image and that fake image will be the output of the generator. 25 00:01:49,710 --> 00:01:52,850 It will be the fake image generated by the generator. 26 00:01:53,040 --> 00:02:00,480 But to get this fake image we need to create this random vector noise that will be exactly this input. 27 00:02:00,480 --> 00:02:04,400 The input of the generator and this input will have a size of 100. 28 00:02:04,590 --> 00:02:09,030 It will become much more clear when we start the training of our two brains. 29 00:02:09,030 --> 00:02:15,300 The brain of the generator and the brain of the discriminator because right now you only see some names 30 00:02:15,300 --> 00:02:22,140 for the variable such as inputs but you'll see that we will create those actual inputs later on in the 31 00:02:22,140 --> 00:02:23,250 training phase. 32 00:02:23,250 --> 00:02:29,310 So no worries if it's not 100 percent crystal clear now it will become much clearer when we start the 33 00:02:29,310 --> 00:02:30,840 training face for now. 34 00:02:30,840 --> 00:02:36,220 Just make sure to understand that we're creating the generator and the discriminator. 35 00:02:36,360 --> 00:02:41,130 And right now we're making the forward function that will forward propagate the signal inside the neural 36 00:02:41,130 --> 00:02:43,440 network at the generator. 37 00:02:43,440 --> 00:02:47,540 All right so now things are going to get super easy. 38 00:02:47,750 --> 00:02:55,680 We are going to take our self duct main memory module which is the metal module of our neural network 39 00:02:55,770 --> 00:02:58,220 object represented by self. 40 00:02:58,270 --> 00:03:06,300 We're going to feed this neural network with the inputs which so far is just an argument but then will 41 00:03:06,300 --> 00:03:14,360 become noise random vector we are fitting the new one that work with the input and this will return 42 00:03:14,660 --> 00:03:22,580 the output because the signal is propagated through all the different modules layers of this neural 43 00:03:22,580 --> 00:03:29,980 network and then in the end we'll get the output which of the three channels of the fake generated images. 44 00:03:30,260 --> 00:03:37,550 And therefore since we get now here I am introducing a new variable that I'm calling output and that 45 00:03:37,550 --> 00:03:43,190 is equal to exactly what is returned by the main metor module. 46 00:03:43,190 --> 00:03:49,550 All right so now that we get the output which just need to return it and therefore I'm just entering 47 00:03:49,910 --> 00:03:53,750 the fourth function with return output. 48 00:03:54,020 --> 00:03:58,400 And there we go we have our output of the generator. 49 00:03:58,400 --> 00:03:59,310 Perfect. 50 00:03:59,330 --> 00:04:06,200 So double congratulations because you made the forward function and because you're done with the whole 51 00:04:06,200 --> 00:04:11,370 class defining the generator but be careful we just defined the class. 52 00:04:11,420 --> 00:04:13,110 Nothing is created yet. 53 00:04:13,160 --> 00:04:19,460 And so in the next to toile we'll have to create an instance that is not checked of the class of the 54 00:04:19,460 --> 00:04:20,570 generator class. 55 00:04:20,690 --> 00:04:26,510 And this object will be nothing else than the neural network of the generator defined by the following 56 00:04:26,510 --> 00:04:33,560 properties and containing the tool or the Ford function to propagate the signal inside the neural network. 57 00:04:33,830 --> 00:04:37,270 So let's create this object in the next tutorial. 58 00:04:37,280 --> 00:04:43,100 It will be quick and easy will just create the object and we'll just apply the weights in it function 59 00:04:43,460 --> 00:04:47,000 to initialize the weights of the neural network as it should be. 60 00:04:47,000 --> 00:04:50,860 That is according to the conventions of the adversarial networks. 61 00:04:50,870 --> 00:04:52,590 So let's do that in the next tutorial. 62 00:04:52,610 --> 00:04:54,490 And until then enjoy computer vision. 7004