|
886 | 886 | }, |
887 | 887 | "cell_type": "markdown", |
888 | 888 | "source": [ |
889 | | - "This model expects it's input to be normalized to the `[-1,1]` range:\n", |
| 889 | + "This model expects its input to be normalized to the `[-1,1]` range:\n", |
890 | 890 | "\n", |
891 | 891 | "```\n", |
892 | 892 | "help(keras_applications.mobilenet_v2.preprocess_input)\n", |
|
907 | 907 | }, |
908 | 908 | "cell_type": "markdown", |
909 | 909 | "source": [ |
910 | | - "So before the passing it to the mobilnet model, we need to convert the input from a range of `[0,1]` to `[-1,1]`." |
| 910 | + "So before the passing it to the MobilNet model, we need to convert the input from a range of `[0,1]` to `[-1,1]`." |
911 | 911 | ] |
912 | 912 | }, |
913 | 913 | { |
|
933 | 933 | }, |
934 | 934 | "cell_type": "markdown", |
935 | 935 | "source": [ |
936 | | - "The mobilenet returns a `6x6` spatial grid of features for each image.\n", |
| 936 | + "The MobileNet returns a `6x6` spatial grid of features for each image.\n", |
937 | 937 | "\n", |
938 | 938 | "Pass it a batch of images to see:" |
939 | 939 | ] |
|
973 | 973 | }, |
974 | 974 | "cell_type": "markdown", |
975 | 975 | "source": [ |
976 | | - "So build a model wrapped around mobilenet, and use `tf.keras.layers.GlobalAveragePooling2D` to average over those space dimensions, before the output `tf.keras.layers.Dense` layer:" |
| 976 | + "So build a model wrapped around MobileNet, and use `tf.keras.layers.GlobalAveragePooling2D` to average over those space dimensions, before the output `tf.keras.layers.Dense` layer:" |
977 | 977 | ] |
978 | 978 | }, |
979 | 979 | { |
|
1371 | 1371 | }, |
1372 | 1372 | "cell_type": "markdown", |
1373 | 1373 | "source": [ |
1374 | | - "#### Image Strings\n", |
| 1374 | + "#### Raw image data\n", |
1375 | 1375 | "\n", |
1376 | | - "TFRecord files are a simple format to store a sequence of binary blobs. In this case the fact that these are read linearly off disk, which may give some performance boost.\n", |
| 1376 | + "TFRecord files are a simple format to store a sequence of binary blobs. By packing multiple examples into the same file, TensorFlow is able to read multiple examples at once, which is especially important for performance when using a remote storage service such as GCS.\n", |
1377 | 1377 | "\n", |
1378 | | - "It also gives you a portable data file that is apropriate, for example, for streaming over a network.\n", |
1379 | | - "\n", |
1380 | | - "First, build a `TFRecord` file from the raw image data:" |
| 1378 | + "First, build a TFRecord file from the raw image data:" |
1381 | 1379 | ] |
1382 | 1380 | }, |
1383 | 1381 | { |
|
0 commit comments