Python 让它更快地解决问题？

我在i7第8代上运行这个。但是与处理器相比，它看起来太慢了

有什么方法可以提高脚本的性能吗？平均CPU使用率仅为10-20%。所以我在寻求你们的帮助。如果你有任何建议，请不要犹豫。

---

=======================================================================================================

在评论讨论之后，20~30%可能已经是您的CPU的最大处理能力，因为它只能使用一个核心。这是因为Python不同于，例如，它不容易并行化。因此，并行化代码可能会使您受益。考虑到您正在使用tensorflow，这可能会让您感兴趣。

@ThomasSablik如果CPU使用率为10-20%，那么瓶颈显然不是编程语言，而是I/O。@Selcuk您在说什么I/O？在逻辑部分我看不到任何I/O。我认为20~30%已经是你核心的100%。要运行得更快，您需要对其进行并行化。也许检查Python可能需要更多的时间，因为它是解释的，当然，但是autor正在做的大部分繁重的处理工作都是由tensorflow完成的，tensorflow是用C编写的，use numpy也是用C编写的。所以在编译完之后，这不会有多大区别，所以第一步应该是分析代码以发现瓶颈。

import os
import sys
os.environ["CUDA_VISIBLE_DEVICES"] = "0,1"
import tensorflow as tf
import numpy as np

PATH = os.getcwd().replace('\\', '/')
ADD_PATH = PATH + '/models/research'

sys.path.insert(0, ADD_PATH)

from object_detection.utils import label_map_util

PATH_TO_FROZEN_GRAPH = 'frozen_inference_graph.pb'
PATH_TO_LABELS = 'labelmap.pbtxt'
NUM_CLASSES = 37

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

def load_model():
    detection_graph = tf.Graph()
    with detection_graph.as_default():
        od_graph_def = tf.GraphDef()
        with tf.gfile.GFile(PATH_TO_FROZEN_GRAPH, 'rb') as fid:
            serialized_graph = fid.read()
            od_graph_def.ParseFromString(serialized_graph)
            tf.import_graph_def(od_graph_def, name='')
            with detection_graph.as_default():
                config = tf.ConfigProto()
                if os.environ['CUDA_VISIBLE_DEVICES'] == '-1':
                    #cpu
                    print("few")
                    #config.inter_op_parallelism_threads = 4
                    #config.intra_op_parallelism_threads = 4
                else:
                    #gpu
                    print("few")
                    #config.gpu_options.allow_growth = True
                    #config.gpu_options.per_process_gpu_memory_fraction = 1.0

                config.log_device_placement = False
                sess = tf.Session(config=config, graph=detection_graph)
                return sess, detection_graph

def inference(sess, detection_graph, img_arr, average_distance_error=3):
        image_np_expanded = np.expand_dims(img_arr, axis=0)
        # Actual detection.
        image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
        boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
        scores = detection_graph.get_tensor_by_name('detection_scores:0')
        classes = detection_graph.get_tensor_by_name('detection_classes:0')
        num_detections = detection_graph.get_tensor_by_name('num_detections:0')
        # Visualization of the results of a detection.
        (boxes, scores, classes, num_detections) = sess.run(
            [boxes, scores, classes, num_detections],
            feed_dict={image_tensor: image_np_expanded})

        # Bellow we do filtering stuff
        captcha_array = []
        # loop our all detection boxes
        for i,b in enumerate(boxes[0]):
            for Symbol in range(37):
                if classes[0][i] == Symbol: # check if detected class equal to our symbols
                    if scores[0][i] >= 0.30: # do something only if detected score more han 0.65
                                        # x-left        # x-right
                        mid_x = (boxes[0][i][1]+boxes[0][i][3])/2 # find x coordinates center of letter
                        # to captcha_array array save detected Symbol, middle X coordinates and detection percentage
                        captcha_array.append([category_index[Symbol].get('name'), mid_x, scores[0][i]])

        # rearange array acording to X coordinates datected
        for number in range(20):
            for captcha_number in range(len(captcha_array)-1):
                if captcha_array[captcha_number][1] > captcha_array[captcha_number+1][1]:
                    temporary_captcha = captcha_array[captcha_number]
                    captcha_array[captcha_number] = captcha_array[captcha_number+1]
                    captcha_array[captcha_number+1] = temporary_captcha


        # Find average distance between detected symbols
        average = 0
        captcha_len = len(captcha_array)-1
        while captcha_len > 0:
            average += captcha_array[captcha_len][1]- captcha_array[captcha_len-1][1]
            captcha_len -= 1
        # Increase average distance error
        average = average/(len(captcha_array)+average_distance_error)

        
        captcha_array_filtered = list(captcha_array)
        captcha_len = len(captcha_array)-1
        while captcha_len > 0:
            # if average distance is larger than error distance
            if captcha_array[captcha_len][1]- captcha_array[captcha_len-1][1] < average:
                # check which symbol has higher detection percentage
                if captcha_array[captcha_len][2] > captcha_array[captcha_len-1][2]:
                    del captcha_array_filtered[captcha_len-1]
                else:
                    del captcha_array_filtered[captcha_len]
            captcha_len -= 1

        # Get final string from filtered CAPTCHA array
        captcha_string = ""
        for captcha_letter in range(len(captcha_array_filtered)):
            captcha_string += captcha_array_filtered[captcha_letter][0]
        return captcha_string