How do I solve the IndexError: list index out of range error when using tensorflow-gpu==2.5.0 for execute the demo_patch.ipynb program

[s5377689]

from attack_utils.patch_utils import extract_roi, evaluate_vanishing_patch, evaluate_mislabeling_patch
from dataset_utils.preprocessing import letterbox_image_padded
from models.yolov3 import YOLOv3_Darknet53
from misc_utils.visualization import visualize_detections
from tensorflow.keras import backend as K
from tqdm import tqdm
from PIL import Image
import numpy as np
import datetime
import os
K.clear_session()

weights = 'model_weights/YOLOv3_Darknet53.h5' # TODO: Change this path to the victim model's weights
#weights = '/research/projects/robust-object-detection/model_weights/YOLOv3_VOC0712_Darknet53.h5' # TODO: Change this path to the victim model's weights

detector = YOLOv3_Darknet53(weights=weights)

---

WARNING:tensorflow:
The following Variables were used a Lambda layer's call (tf.nn.convolution), but
are not present in its tracked objects:
<tf.Variable 'conv2d/kernel:0' shape=(3, 3, 3, 32) dtype=float32>
It is possible that this is intended behavior, but it is more likely
an omission. This is a strong indication that this layer should be
formulated as a subclassed Layer rather than a Lambda layer.
WARNING:tensorflow:
The following Variables were used a Lambda layer's call (tf.compat.v1.nn.fused_batch_norm), but
are not present in its tracked objects:
<tf.Variable 'batch_normalization/gamma:0' shape=(32,) dtype=float32>
<tf.Variable 'batch_normalization/beta:0' shape=(32,) dtype=float32>
It is possible that this is intended behavior, but it is more likely
an omission. This is a strong indication that this layer should be
formulated as a subclassed Layer rather than a Lambda layer.
WARNING:tensorflow:
The following Variables were used a Lambda layer's call (tf.nn.convolution_1), but
are not present in its tracked objects:
<tf.Variable 'conv2d_1/kernel:0' shape=(3, 3, 32, 64) dtype=float32>
It is possible that this is intended behavior, but it is more likely
an omission. This is a strong indication that this layer should be
formulated as a subclassed Layer rather than a Lambda layer.
WARNING:tensorflow:
The following Variables were used a Lambda layer's call (tf.compat.v1.nn.fused_batch_norm_1), but
are not present in its tracked objects:
...
<tf.Variable 'batch_normalization_58/beta:0' shape=(256,) dtype=float32>
It is possible that this is intended behavior, but it is more likely
an omission. This is a strong indication that this layer should be
formulated as a subclassed Layer rather than a Lambda layer.
Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...

IndexError Traceback (most recent call last)
C:\TEMP\ipykernel_9648\2429568344.py in
2 #weights = '/research/projects/robust-object-detection/model_weights/YOLOv3_VOC0712_Darknet53.h5' # TODO: Change this path to the victim model's weights
3
----> 4 detector = YOLOv3_Darknet53(weights=weights)

c:\Users\User\Desktop\TOG\models\yolov3.py in init(self, weights, model_img_size, confidence_thresh_default, confidence_thresh_eval)
241 def init(self, weights, model_img_size=(416, 416), confidence_thresh_default=0.20, confidence_thresh_eval=0.01):
242 super().init(weights, yolo_darknet53,
--> 243 model_img_size, confidence_thresh_default, confidence_thresh_eval)
244
245

c:\Users\User\Desktop\TOG\models\yolov3.py in init(self, weights, backbone, model_img_size, confidence_thresh_default, confidence_thresh_eval)
18 self.num_anchors = len(self.anchors)
19
---> 20 self.model = backbone(Input(shape=(None, None, 3)), self.num_anchors // 3, self.num_classes)
21 self.model.load_weights(weights)
22

c:\Users\User\Desktop\TOG\yolov3_utils\model.py in yolo_darknet53(inputs, num_anchors, num_classes)
69
70 x = compose(DarknetConv2D_BN_Leaky(256, (1, 1)), UpSampling2D(2))(x)
---> 71 x = Concatenate()([x, darknet.layers[152].output])
72 x, y2 = make_last_layers(x, 256, num_anchors * (num_classes + 5))
73 y2 = Reshape((tf.shape(y2)[1], tf.shape(y2)[2], num_anchors, num_classes + 5))(y2)

IndexError: list index out of range

TOG_yolov3_utils_model.txt

[s5377689]

從attack_utils.patch_utils導入extract_roi，evaluate_vanishing_patch，evaluate_mislabeling_patch 從dataset_utils.preprocessing導入letterbox_image_padded 從models.yolov3導入YOLOv3_Darknet53 從misc_utils.visualization導入visible_detections 從tensorflow.keras導入後端作為K 從tqdm導入tqdm 從PIL導入圖像 導入numpy作為np 導入日期時間 導入 os K.clear_session()

weights = 'model_weights/YOLOv3_Darknet53.h5' # TODO：將此路徑更改為受害者模型的權重 #weights = '/research/projects/robust-object-detection/model_weights/YOLOv3_VOC0712_Darknet53.h5' # TODO：將此路徑更改為受害者模型的權重

檢測器= YOLOv3_Darknet53（權重=權重）

警告：tensorflow：

以下變量用於 Lambda 層的調用 (tf.nn.volving)，但
不存在於其跟踪對像中：
<tf.Variable 'conv2d/kernel:0' shape=(3, 3, 3, 32) dtype=float32>
這可能是有意的行為，但更可能是
遺漏。這強烈表明該層應被
制定為子類層而不是 Lambda 層。
警告：tensorflow：
以下變量用於 Lambda 層的調用 (tf.compat.v1.nn.fused_batch_norm)，但
不存在於其跟踪對像中：
<tf.Variable 'batch_normalization/gamma:0' shape=(32, ) dtype=float32>
<tf.Variable 'batch_normalization/beta:0' shape=(32,
這可能是有意的行為，但更有可能是
疏忽。這強烈表明該層應被
制定為子類層而不是 Lambda 層。
警告：tensorflow：
以下變量用於 Lambda 層的調用 (tf.nn.convolution_1)，但
不存在於其跟踪對像中：
<tf.Variable 'conv2d_1/kernel:0' shape=(3, 3, 32, 64) dtype=float32>
這可能是有意的行為，但更可能是
遺漏。這強烈表明該層應被
制定為子類層而不是 Lambda 層。
警告：張量流：
以下變量用於 Lambda 層的調用 (tf.compat.v1.nn.fused_batch_norm_1)，但
不存在於其跟踪對像中：
...
<tf.Variable 'batch_normalization_58/beta:0' shape=(256,) dtype=float32>
這可能是有意的行為，但更可能是
疏忽。這強烈表明該層應被
制定為子類層而不是 Lambda 層。
輸出被截斷。作為可滾動元素查看或在文本編輯器中打開。調整單元格輸出設置...
IndexError Traceback (most recent call last) C:\TEMP\ipykernel_9648\2429568344.py in 2 #weights = '/research/projects/robust-object-detection/model_weights/YOLOv3_VOC0712_Darknet53.h5' # TODO: Change this path to the victim model's weights 3 ----> 4 detector = YOLOv3_Darknet53(weights=weights)

c:\Users\User\Desktop\TOG\models\yolov3.py in init(self, weights, model_img_size, confidence_thresh_default, confidence_thresh_eval) 241 def init(self, weights, model_img_size=(416, 416), confidence_thresh_default=0.20, confidence_thresh_eval=0.01): 242 super().init(weights, yolo_darknet53, --> 243 model_img_size, confidence_thresh_default, confidence_thresh_eval) 244 245

c:\Users\User\Desktop\TOG\models\yolov3.py in init(self, weights, backbone, model_img_size, confidence_thresh_default, confidence_thresh_eval) 18 self.num_anchors = len(self.anchors) 19 ---> 20 self.model = backbone(Input(shape=(None, None, 3)), self.num_anchors // 3, self.num_classes) 21 self.model.load_weights(weights) 22

c:\Users\User\Desktop\TOG\yolov3_utils\model.py in yolo_darknet53(inputs, num_anchors, num_classes) 69 70 x = compose(DarknetConv2D_BN_Leaky(256, (1, 1)), UpSampling2D(2))(x) ---> 71 x = Concatenate()([x, darknet.layers[152].output]) 72 x, y2 = make_last_layers(x, 256, num_anchors * (num_classes + 5)) 73 y2 = Reshape((tf) .shape(y2)[1], tf.shape(y2)[2], num_anchors, num_classes + 5))(y2)

IndexError：列表索引超出範圍

TOG_yolov3_utils_model.txt