-
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcheck_images.txt
51 lines (36 loc) · 2.53 KB
/
check_images.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Questions regarding Uploaded Image Classification:
1. Did the three model architectures classify the breed of dog in Dog_01.jpg to be the same breed? If not, report the differences in the classifications.
Answer:
vgg:
Dog_01.jpg:
Real: dog Classifier: golden retriever
resnet: same
alexnet: same
The three model architectures classify the breed of dog in Dog_01.jpg to be the same breed.
2. Did each of the three model architectures classify the breed of dog in Dog_01.jpg to be the same breed of dog as that model architecture classified Dog_02.jpg? If not, report the differences in the classifications.
Answer:
vgg classify the breed of dog in Dog_01.jpg is the same breed of dog as Dog_02.jpg.
Dog_01.jpg:
Real: dog Classifier: golden retriever
Dog_02.jpg:
Real: dog Classifier: golden retriever
resnet and alexnet classify the breed of dog in Dog_01.jpg not the same breed of dog as Dog_02.jpg.
Dog_01.jpg:
Real: dog Classifier: golden retriever
Dog_02.jpg:
Real: dog Classifier: cocker spaniel, english cocker spaniel, cocker
3. Did the three model architectures correctly classify Animal_Name_01.jpg and Object_Name_01.jpg to not be dogs? If not, report the misclassifications.
Answer: The three model architectures correctly classified Animal_Name_01.jpg and Object_Name_01.jpg to not be dogs.
Object_Name_01.jpg:
Real: object name Classifier: folding chair
PetLabelDog: 0 ClassLabelDog: 0
Animal_Name_01.jpg:
Real: animal name Classifier: egyptian cat, cat
PetLabelDog: 0 ClassLabelDog: 0
4. Based upon your answers for questions 1. - 3. above, select the model architecture that you feel did the best at classifying the four uploaded images. Describe why you selected that model architecture as the best on uploaded image classification.
Answer:
I think the model architecture that does the best job of classifying the four uploaded images is vgg.
Compared to ResNet and AlexNet, VGG provides better accuracy and performance.
Although ResNet and AlexNet were slightly faster, they classified Dog_02.jpg as the wrong breed,
while the VGG architecture was still able to classify Dog_02.jpg as golden retriever.
VGG provides the best performance in terms of accuracy, although it may be slower than the other two architectures.