WEBVTT 00:00.080 --> 00:00.960 Welcome back. 00:01.000 --> 00:07.680 In order to save time, we're going to update the codes of the previous section. 00:07.680 --> 00:15.000 So I created this project Object Detection SSD app in the previous section. 00:15.160 --> 00:24.760 Go and open it in order to update some codes and add the model to the ML package and run it. 00:24.800 --> 00:25.240 Okay. 00:25.280 --> 00:32.240 It's similar to what we've done before, but we need to change and the codes according to the model 00:32.240 --> 00:32.960 shapes. 00:33.200 --> 00:33.720 Okay. 00:33.920 --> 00:40.040 So the manifest we should create the the permission of the camera. 00:40.160 --> 00:45.280 Also you can create your own um, and new uh, project. 00:45.400 --> 00:54.560 Uh, but you should include the, those codes, the permissions for the manifest inside Gradle we should 00:54.600 --> 00:56.720 include TensorFlow. 00:56.760 --> 01:01.040 We should include Camerax permission and hilt. 01:01.080 --> 01:07.680 We talked about hilt in the previous sections and in the previous videos. 01:07.680 --> 01:13.160 So we need to add those, um, those plugins and also add the project level. 01:13.160 --> 01:15.120 We should add those okay. 01:15.400 --> 01:23.080 So those are the permissions and the libraries needed to work with our application. 01:23.080 --> 01:29.640 Now delete this ML package and delete the Coco data set. 01:29.680 --> 01:34.320 Now let's add a new model that we created. 01:34.320 --> 01:34.880 Before. 01:34.920 --> 01:44.240 Go to the downloads, copy the best Float16 the Tflite to the ML and here we go. 01:44.280 --> 01:50.480 Now create a new file, name it as labels dot txt. 01:50.520 --> 01:55.640 Inside this label we need to add the object classes. 01:55.640 --> 02:02.240 So all the classes that you annotate in Roboflow you should add it here. 02:02.360 --> 02:05.120 Okay, we have only one. 02:05.400 --> 02:09.480 Um, we have only one class which is the battery. 02:09.480 --> 02:17.200 So if you remember, the class is named battery and you should write it exactly as you specified in 02:17.200 --> 02:21.720 Roboflow and in your, um, in your data set. 02:21.760 --> 02:22.200 Okay. 02:22.240 --> 02:29.760 So this is the labels dot txt and this is the best Float16 Tflite this those are the three steps we've 02:29.760 --> 02:37.760 done before the manifest, the Build.gradle and the best Float16 Tflite and labels. 02:37.800 --> 02:41.720 Now let's start examine every uh, every class. 02:41.720 --> 02:46.320 So here the response inside the util we have this class. 02:46.360 --> 02:51.760 It's, uh, we don't need to make anything uh, any update to it. 02:51.800 --> 02:59.440 We have loading success and failure as we told the talked about it in the previous videos, this class 02:59.440 --> 03:08.360 represents the possible states of a synchronous operation or API response out of T. 03:08.400 --> 03:15.810 This is the type T indicates that that type parameter t is covenant. 03:15.850 --> 03:20.690 You can use response of dog where response animal is expected. 03:20.690 --> 03:30.410 So this is can be this can be uh like understood in this way response of dog where response animal is 03:30.410 --> 03:31.330 expected. 03:31.370 --> 03:31.850 Okay. 03:31.890 --> 03:35.530 So this is the type T detection object. 03:35.730 --> 03:41.250 We have this detection objects as uh similar to what we've done before. 03:41.490 --> 03:47.970 It's a data class contains the score label and the bounding box. 03:48.090 --> 03:53.170 The detection parameters also similar to what we've done before. 03:53.450 --> 03:59.210 So those three classes, we don't need to make any change inside them. 03:59.250 --> 04:02.770 Also I'll skip the object detector for now. 04:03.010 --> 04:04.770 Camera permission wrapper. 04:04.970 --> 04:08.290 There's nothing to do with this file. 04:08.330 --> 04:09.730 Camera preview. 04:10.090 --> 04:12.250 I'll skip it for later. 04:12.330 --> 04:15.050 Uh, edit the detection screen. 04:15.090 --> 04:16.970 No need to do anything. 04:17.170 --> 04:21.370 Of course the detection view model we're going to update it. 04:21.410 --> 04:22.530 The detection app. 04:22.530 --> 04:26.890 No need to uh to do anything any update. 04:26.930 --> 04:28.170 The main activity. 04:28.330 --> 04:29.930 No need to do anything. 04:29.930 --> 04:35.130 So we're going to update the detection view model. 04:35.330 --> 04:39.010 We're going to update the object detector. 04:39.130 --> 04:43.050 And if there is any error we're going to update the corresponding files. 04:43.050 --> 04:47.170 But for now we uh added the permission. 04:47.370 --> 04:54.490 We added that Gradle libraries and the model best Float16 Tflite. 04:54.490 --> 04:58.850 And the label and those classes are similar. 04:58.890 --> 05:06.210 The response detection, object detection parameter, camera permission wrapper, camera preview, the 05:06.530 --> 05:10.290 drill boxes, detection app and main activity. 05:10.450 --> 05:16.490 We're going to update the object detector detection screen and detection view model.