我使用 Python 和 Keras 制作了一個(gè)卷積神經(jīng)網(wǎng)絡(luò)。我在測(cè)試集上測(cè)試我的模型，每個(gè)類的圖像數(shù)量是隨機(jī)的（1 個(gè)文件夾包含 x 數(shù)量的圖像）。我能夠獲得一個(gè)數(shù)據(jù)框，其中顯示圖像和目錄的文件名以及預(yù)測(cè)。我想從文件名中刪除目錄。它隨機(jī)顯示 350 張圖像/dogs1.tif，我希望它只顯示 dogs1.tif。  #import my modelnew_model = tf.keras.models.load_model('model folder')#upload my test datatrain_datagen = ImageDataGenerator(rescale=1./255)test_batches = train_datagen.flow_from_directory(        'folder containing random images',        target_size=(224, 224),        batch_size=10,        classes = None,        class_mode = None,        shuffle = False)#my predictionpredictions = new_model.predict(test_batches, steps=35, verbose=0)#rounding my predctionsrounded_predictions = np.argmax(predictions, axis = -1)#converting one hot encoded labels to categorical labels labels =["dog","cat","horse"]names = [0,1,2]labels_name = dict(zip(names, labels))#joining them togetherlabels_name = dict((v,k) for k,v in labels_name.items())predictions = [labels[k] for k in rounded_predictions]#getting files names for the imagesfilenames= test_batches.filenames#creating the dataframeresults=pd.DataFrame({"file":filenames,"pr":predictions})