Catal Vision API Tests
Intro
0.1
Comments, ideas in random order
1
Some general introspection
1.1
Frequency of all labels
1.2
Frequency of unique labels
1.3
Distribution of scores
1.4
Top ranking labels (on average)
1.5
Score distribution for those top ranking labels
1.6
Most used labels (more than 50 times) and their mean scores
1.7
Highest scoring labels (GV over 0.8, CP over .95) and their frequency
1.8
Labels occuring in both sets: GV and CP
2
OCR images with extracted text
3
CP Label: “adult”
4
CP Label: “architecture”
5
CP Label: “building”
6
CP Label: “cave”
7
CP Label: “cement”
8
CP Label: “dirty”
9
CP Label: “geology”
10
CP Label: “group”
11
CP Label: “man”
12
CP Label: “no person”
13
CP Label: “people”
14
CP Label: “rock”
15
CP Label: “sand”
16
CP Label: “soil”
17
CP Label: “stone”
18
CP Label: “texture”
19
CP Label: “travel”
20
CP Label: “wall”
21
CP Label: “woman”
22
GV Entity: “Archaeology”
23
GV Entity: “Çatalhöyük”
24
GV Entity: “Laborer”
25
GV Label: “ancient history”
26
GV Label: “archaeology”
27
GV Label: “archaeological site”
28
GV Label: “artifact”
29
GV Label: “building”
30
GV Label: “fauna”
31
GV Label: “geology”
32
GV Label: “rock”
33
GV Label: “ruins”
34
GV Label: “sand”
35
GV Label: “social group”
36
GV Label: “soil”
37
GV Label: “structure”
38
GV Label: “wall”
39
CP AND GV Label: “archaeology”
40
CP AND GV Label: “building”
41
CP AND GV Label: “geology”
42
CP AND GV Label: “sand”
43
CP AND GV Label: “soil”
44
CP AND GV Label: “wall”
45
Images 1-100
46
Images 101-200
47
Images 201-300
48
Images 301-400
49
Images 401-500
50
Images 501-600
51
Images 601-700
52
Images 701-766
Published with bookdown
Catal Image Vision API Tests
Chapter 8
CP Label: “dirty”
Table 8.1:
‘dirty’ - confidence score > 0.95
col1
col2
