6bi&ASrSSKrSSKrSSKrSSKJr SSKJr SSKJ r SSK J r Sq Sr SrS rS rS rS rS rSr\ "S5SSj5r\R-\S\S9\l\ "S5SSj5rSrSrSSjrSrSrg)z@Utilities for ImageNet data preprocessing & prediction decoding.N) activations)backend) data_utils) keras_exportzUhttps://storage.googleapis.com/download.tensorflow.org/data/imagenet_class_index.jsonam Preprocesses a tensor or Numpy array encoding a batch of images. Usage example with `applications.MobileNet`: ```python i = tf.keras.layers.Input([None, None, 3], dtype = tf.uint8) x = tf.cast(i, tf.float32) x = tf.keras.applications.mobilenet.preprocess_input(x) core = tf.keras.applications.MobileNet() x = core(x) model = tf.keras.Model(inputs=[i], outputs=[x]) image = tf.image.decode_png(tf.io.read_file('file.png')) result = model(image) ``` Args: x: A floating point `numpy.array` or a `tf.Tensor`, 3D or 4D with 3 color channels, with values in the range [0, 255]. The preprocessed data are written over the input data if the data types are compatible. To avoid this behaviour, `numpy.copy(x)` can be used. data_format: Optional data format of the image tensor/array. None, means the global setting `tf.keras.backend.image_data_format()` is used (unless you changed it, it uses "channels_last").{mode} Defaults to `None`. Returns: Preprocessed `numpy.array` or a `tf.Tensor` with type `float32`. {ret} Raises: {error} a mode: One of "caffe", "tf" or "torch". - caffe: will convert the images from RGB to BGR, then will zero-center each color channel with respect to the ImageNet dataset, without scaling. - tf: will scale pixels between -1 and 1, sample-wise. - torch: will scale pixels between 0 and 1 and then will normalize each channel with respect to the ImageNet dataset. Defaults to "caffe". zE ValueError: In case of unknown `mode` or `data_format` argument.z; ValueError: In case of unknown `data_format` argument.zH The inputs pixel values are scaled between -1 and 1, sample-wise.z The input pixels values are scaled between 0 and 1 and each channel is normalized with respect to the ImageNet dataset.z The images are converted from RGB to BGR, then each color channel is zero-centered with respect to the ImageNet dataset, without scaling.z2keras.applications.imagenet_utils.preprocess_inputcUS;a[SU35eUc[R"5nOUS;a[SU35e[U[R 5(a [ XUS9$[XUS9$)z@Preprocesses a tensor or Numpy array encoding a batch of images.>tfcaffetorchzDExpected mode to be one of `caffe`, `tf` or `torch`. Received: mode=> channels_lastchannels_firstz]Expected data_format to be one of `channels_first` or `channels_last`. Received: data_format=) data_formatmode) ValueErrorrimage_data_format isinstancenpndarray_preprocess_numpy_input_preprocess_symbolic_input)xr rs b/home/james-whalen/.local/lib/python3.13/site-packages/tf_keras/src/applications/imagenet_utils.pypreprocess_inputrhs ++ "V %  //1 ? ? 66A] D  !RZZ  &qMM)!4PP)rreterrorz4keras.applications.imagenet_utils.decode_predictionsc T[UR5S:wdURSS:wa![S[UR5-5e[cE[ R "S[SSS 9n[U5n[R"U5qSSS5 /nUHpnUR5U*SSSS 2nUVs/sH$n[[[U55XW4-PM& nnURS S S 9 URU5 Mr U$!,(df  N=fs snf)aDecodes the prediction of an ImageNet model. Args: preds: Numpy array encoding a batch of predictions. top: Integer, how many top-guesses to return. Defaults to 5. Returns: A list of lists of top class prediction tuples `(class_name, class_description, score)`. One list of tuples per sample in batch input. Raises: ValueError: In case of invalid shape of the `pred` array (must be 2D). izx`decode_predictions` expects a batch of predictions (i.e. a 2D array of shape (samples, 1000)). Found array with shape: Nzimagenet_class_index.jsonmodels c2c37ea517e94d9795004a39431a14cb) cache_subdir file_hashc US$)Nr)rs r$decode_predictions..s!A$rT)keyreverse)lenshaperstr CLASS_INDEXrget_fileCLASS_INDEX_PATHopenjsonloadargsorttuplesortappend) predstopfpathfresultspred top_indicesiresults rdecode_predictionsrAs& 5;;1 A$ 6 '*-U[[)9 :  ## ' !8   %[A))A,KGllncTU+DbD1 CNO;a% CF+,z9;O  5v  N[ Ps;D?+D% D"c[URR[R5(d#UR [ R"5SS9nUS:Xa US-nUS-nU$US:XaUS-n/SQn/S QnO?US :Xa)URS :Xa US S S 2S4nOUS S 2S S S 2S4nO USS S S 24n/SQnS nUS :XGa@URS :XaUSS S 2S S 24==US-ss'USS S 2S S 24==US-ss'USS S 2S S 24==US-ss'UbEUSS S 2S S 24==US-ss'USS S 2S S 24==US-ss'USS S 2S S 24==US-ss'U$US S 2SS S 2S S 24==US-ss'US S 2SS S 2S S 24==US-ss'US S 2SS S 2S S 24==US-ss'UbNUS S 2SS S 2S S 24==US-ss'US S 2SS S 2S S 24==US-ss'US S 2SS S 2S S 24==US-ss'U$US==US-ss'US==US-ss'US==US-ss'Ub0US==US-ss'US==US-ss'US==US-ss'U$)aUPreprocesses a Numpy array encoding a batch of images. Args: x: Input array, 3D or 4D. data_format: Data format of the image array. mode: One of "caffe", "tf" or "torch". - caffe: will convert the images from RGB to BGR, then will zero-center each color channel with respect to the ImageNet dataset, without scaling. - tf: will scale pixels between -1 and 1, sample-wise. - torch: will scale pixels between 0 and 1 and then will normalize each channel with respect to the ImageNet dataset. Returns: Preprocessed Numpy array. F)copyr_@?r o@g ףp= ?gv/?gCl?gZd;O?gy&1?g?r Nr$.gjtY@g`"1]@gQ^@rrr).r).r).r) issubclassdtypetyperfloatingastyperfloatxndim)rr rmeanstds rrrs( aggllBKK 0 0 HHW^^%EH 2 t| U  S  U $# * *vv{ddCiLa2slO#tt) A)&& 66Q; aAgJ$q' !J aAgJ$q' !J aAgJ$q' !J!Q' c!f$ !Q' c!f$ !Q' c!f$ " H aAqjMT!W $M aAqjMT!W $M aAqjMT!W $M!Q1* Q' !Q1* Q' !Q1* Q'  H & T!W  & T!W  & T!W ? fIQ I fIQ I fIQ I Hrc US:Xa US-nUS-nU$US:XaUS-n/SQn/SQnOIUS:Xa3[R"U5S :Xa US S S 2S 4nOUS S 2S S S 2S 4nO US S S S 24n/S QnS n[R"[R"U5*5n[R "U5[R "U5:wa@[R "U[R"U[R "U55US9nO[R "XU5nUb^[R"[R"U5[R "U5S9nUS:Xa[R"US5nX-nU$)aMPreprocesses a tensor encoding a batch of images. Args: x: Input tensor, 3D or 4D. data_format: Data format of the image tensor. mode: One of "caffe", "tf" or "torch". - caffe: will convert the images from RGB to BGR, then will zero-center each color channel with respect to the ImageNet dataset, without scaling. - tf: will scale pixels between -1 and 1, sample-wise. - torch: will scale pixels between 0 and 1 and then will normalize each channel with respect to the ImageNet dataset. Returns: Preprocessed tensor. rrDrEr rFrGrHr rINr$.rJ)r )rL)r$rr) rrQconstantrarrayrLbias_addcastreshape)rr rrRrS mean_tensor std_tensors rrrsZ( t| U  S  U $# * *||A!#ddCiLa2slO#tt) A)""BHHTN?3K}}Q7==55    LLgmmA&6 7#    Q[ 9 %%bhhsm7==;KL * * Z@J  HrcUS:waU(a[U5S:XavUS:Xa8USS;a'[R"S[US5-S-SS 9 USX4nOGUS S;a'[R"S[US 5-S-SS 9 XUS 4nOUS:XaSX4nOXS4nUS:Xa"U(aUbX:wa[ S US U35eU$U(aUS:XanUbj[U5S:wa [ S5eUSS:waUS:Xa[ SUS35eUSb USU:dUSbUSU:a[ SUSUSU35eOUbj[U5S:wa [ S5eUS S:waUS:Xa[ SUS35eUSb USU:dUSbUSU:a[ SUSUSU35eOU(aUnO US:XaSnOSnU(aS U;a[ SU35eU$)aInternal utility to compute/validate a model's input shape. Args: input_shape: Either None (will return the default network input shape), or a user-provided shape to be validated. default_size: Default input width/height for the model. min_size: Minimum input width/height accepted by the model. data_format: Image data format to use. require_flatten: Whether the model is expected to be linked to a classifier via a Flatten layer. weights: One of `None` (random initialization) or 'imagenet' (pre-training on ImageNet). If weights='imagenet' input channels must be equal to 3. Returns: An integer shape tuple (may include None entries). Raises: ValueError: In case of invalid argument values. imagenetrIr r>rrIz]This model usually expects 1 or 3 input channels. However, it was passed an input_shape with z input channels.r) stacklevelr$NzXWhen setting `include_top=True` and loading `imagenet` weights, `input_shape` should be z. Received: input_shape=z0`input_shape` must be a tuple of three integers.z6The input must have 3 channels; Received `input_shape=`rzInput size must be at least rz; Received: input_shape=)rINN)NNrIz[If `include_top` is True, you should specify a static `input_shape`. Received: input_shape=)r+warningswarnr-r) input_shape default_sizemin_sizer require_flattenweights default_shapes robtain_input_shaperh7s8*[1AQ1F * *1~V+ B+a.)*)) ! )^\HM2f, B+b/*+)) ! *RIM * * ;M);M*  "+ //--8M;  * *&{#q($Jq>Q&7j+@$((3}A7  N.;q>H3L!!n0[^h5N$6xjH:&''2m5 &{#q($Jr?a'Gz,A$((3}A7  N.;q>H3L!!n0[^h5N$6#*AhZ0''2m5 'K..- -  ; ))4 7  rc@[R"5S:XaSOSn[R"U5X"S-n[U[5(aX4nUScSnOSUSS-- SUSS-- 4nUSS-USS-4nUSUS- US4USUS- US44$)zReturns a tuple for zero-padding for 2D convolution with downsampling. Args: inputs: Input tensor. kernel_size: An integer or tuple/list of 2 integers. Returns: A tuple. r rrr)rr)rr int_shaperint)inputs kernel_sizeimg_dim input_sizeadjustcorrects r correct_padrrs,,.2BBaG""6*7kCJ+s##"0 !}jma''Z]Q->)>?1~"KNa$78G fQi , fQi , rcUcg[R"U5nU[R"S5[R"S51;a[SU35eg)a$validates that the classifer_activation is compatible with the weights. Args: classifier_activation: str or callable activation function weights: The pretrained weights to load. Raises: ValueError: if an activation other than `None` or `softmax` are used with pretrained weights. NsoftmaxzOnly `None` and `softmax` activations are allowed for the `classifier_activation` argument when using pretrained weights, with `include_top=True`; Received: classifier_activation=)rgetr)classifier_activationrfs rvalidate_activationrwsg'OO,AB "% %&;$; =  r)Nr ))N)__doc__r2r`numpyr tf_keras.srcrrtf_keras.src.utilsr tensorflow.python.util.tf_exportrr.r0PREPROCESS_INPUT_DOCPREPROCESS_INPUT_MODE_DOC"PREPROCESS_INPUT_DEFAULT_ERROR_DOCPREPROCESS_INPUT_ERROR_DOCPREPROCESS_INPUT_RET_DOC_TFPREPROCESS_INPUT_RET_DOC_TORCHPREPROCESS_INPUT_RET_DOC_CAFFErformatrArrrhrrrwr&rrrsG $ ): % "H &H">K":"N BCQDQ,066 " ,7DE(F(VF R9 D  vr2 r