Tagger API
API Key
TAGGER API is a REST API that uses HTTP client which requires an API key to use. One can check API key in the Account info page by clicking the 'Show' button.
Basic Information
TAGGER API request URI and its respective required header information are as follows.
Method
Request URI
Required Headers
POST
x-api-key: API Key for authentication [Required]
accept-language: selected language
en : English (Default) ko: Korean
zh: Chinese
ja: Japanese
Request
[HTTP Request Header]
POST /tagger/v2.12/tags HTTP/1.1 Host : api.omnious.com Content-Type: application/json x-api-key : {API Key provided by OMNIOUS} accept-language : {Selected language}
[HTTP Request Body]
Field Name
Data Type
Description
Required
image
object
Object containing type and contents of an image
O
image.type
string
Type of the requested image (urlor base64)
O
image.content
string
Information about the image type
If type is url: URL of the requested image
If type is base64: string of the image encoded in base64
A prefix is required for the base64 string. Please see below the chart for details.
O
context
object
Parent object containing additional information
X
context.id
string
Product id as kept by the user
X
context.detection
string array
value : TOP, BOTTOM, WHOLEBODY, SWIMWEAR, SHOES
If only shoes are to be tagged
detection:["SHOES"]
If only skirts and pants are to be tagged
detection: ["BOTTOM"]
If all detected items (bbox) are to be tagged [default]
detection: undefined or null or []
(Since tagging for shoes are provided as a beta service, the default value is 'CLOTHING', so if tagging for shoes are required, 'SHOES' must be given explicitly.)
X
context.gender
string
Gender information about the product
example) female, male , unisex
X
context.brand
string
Product's brand name
X
context.currency
string
Currency
Example) KRW or USD
X
context.price
string
Product price
X
context.age
string
The target age group of the product
Example) 10s, early20s, mid-late20s, early30s, mid-late30s, 40s, over50s
X
context.keywords
string
Product keywords
Example) summer onepiece
X
context.productContentName
string
Product name of the requested image
X
context.season
string
Season information about the product
Example) 19SS, 19FW, 20SS, 20FW
X
base64 python code example
encoded = base64.b64encode(image_data).decode('ascii')
img2str'data:image/png;base64,{}'.format(encoded)Example request
{
"image":{
"type":"url",
"content":"Image URL"
},
"context":{
"id": "A111",
"detection": ["TOP", "BOTTOM", "WHOLEBODY", "SWIMWEAR", "SHOES"],
"gender": "male",
"brand": "NIKE",
"currency":"USD",
"price": 299.99,
"age": "early20s",
"keywords": "bling bling",
"productContentName": "one-piece",
"season": "19SS" }
}Response
Analysis performed by the Tagger API is returned as a JSON format. Descriptions of each field are as follows.
Field Name
Data Type
Description
status
string
Whether image analysis is successfully completed.
data
object
Object containing image analysis results
data.imageinfo[]
object array
Array containing information on inserted image
data.imageinfo[].id
string
Image information id
data.imageinfo[].name
string
Image information description
data.genderInfo
object
Image information description
data.genderInfo.id
string
Object containing gender information, if a model was detected from the image.
data.genderInfo.name
string
male: when all models in the image are male
female: when all models in the image are female
unknown: when the genders are either mixed, ambiguous, or if no models were detected.
data.objects[]
object array
Array containing the attribute information for each detected fashion items in the given image
data.objects[].type
string
“CLOTHING”, “SWIMWEAR”, ""BAGS“, "HATS", "JEWELRIES", "ACCESSORIES", "NODETC” Provided in one of these three types
data.objects[].tags[]
object array
Array of objects containing tag information for each detected fashion item.
Below data is provided per detected type and its respective bounding boxes.
Empty array if no type is detected.
data.objects[].tags[].category
object
Object containing the category information
data.objects[].tags[].category.id
string
Category ID (Ex: T0001)
data.objects[].tags[].category.name
string
Category description (Example: dress)
data.objects[].tags[].position
object
Object containing the information of the center position of the bounding box.
data.objects[].tags[].position.x
number
x position of the center position of the bounding box Coordinates (Ratio proportional to image size)
data.objects[].tags[].position.y
number
y position of the center position of the bounding box Coordinates (Ratio proportional to image size)
data.objects[].tags[].pairIndex
number
Provided only when data.objects[].type is “SHOES”, which represents the index of the identical product. If more than one shoe is detected in an image, each shoe is given a “pairIndex” value between 0 ~ 100. If Tagger discovers that two shoes are the same, pairIndex value will be same also.
data.objects[].tags[].colors[]
object array
Array of objects containing color information
Each object contains one of the 18 provided colors
Array contains top1, top2, top3 model results.
data.objects[].tags[].colors[].id
string
Color unique ID
data.objects[].tags[].colors[].name
string
Color description
data.objects[].tags[].colors[].confidence
number
Color prediction confidence value (0~1)
Higher the confidence the closer the value is to 1
data.objects[].tags[].colorDetails[]
object array
Array of objects containing color detail information
Provides color detail information from a palette of 2000+ colors.
data.objects[].tags[].colorDetails[].code
string
Color detail HEX code
data.objects[].tags[].colorDetails[].name
string
Color detail description
data.objects[].tags[].colorDetails[].pixelFraction
number
Color detail pixelFraction
data.objects[].tags[].item
object
Object containing the item information
data.objects[].tags[].item.id
string
Item unique ID
data.objects[].tags[].item.name
string
Item description
data.objects[].tags[].item.confidence
number
Item prediction confidence value (0~1) Higher the confidence the closer the value is to 1
data.objects[].tags[].details[]
object array
Array of objects containing detail information
data.objects[].tags[].details[].id
string
Detail unique ID
data.objects[].tags[].details[].name
string
Detail description
data.objects[].tags[].details[].confidence
number
Detail prediction confidence value (0~1) Higher the confidence the closer the value is to 1
data.objects[].tags[].prints[]
object array
Array of objects containing print information
data.objects[].tags[].prints[].id
string
Print unique ID
data.objects[].tags[].prints[].name
string
Print description
data.objects[].tags[].prints[].confidence
number
Print prediction confidence value (0~1) Higher the confidence the closer the value is to 1
data.objects[].tags[].textures[]
object array
Array of objects containing texture information
data.objects[].tags[].textures[].id
string
Texture unique ID
data.objects[].tags[].textures[].name
string
Texture description
data.objects[].tags[].textures[].confidence
number
Texture prediction confidence value (0~1)
Higher the confidence the closer the value is to 1
data.objects[].tags[].looks[]
object array
Array of objects containing look information
data.objects[].tags[].looks[].id
string
Look unique ID
data.objects[].tags[].looks[].name
string
Look description
data.objects[].tags[].looks[].confidence
number
Look prediction confidence value (0~1)
Higher the confidence the closer the value is to 1
data.objects[].tags[].length
object
Object containing length information
data.objects[].tags[].length.id
string
Length unique ID
data.objects[].tags[].length.name
string
Length description
data.objects[].tags[].length.confidence
number
Length prediction confidence value (0~1)
Higher the confidence the closer the value is to 1
data.objects[].tags[].sleeveLength
object
Object containing sleeve length information
data.objects[].tags[].sleeveLength.id
string
Sleeve length unique ID
data.objects[].tags[].sleeveLength.name
string
Sleeve length name
data.objects[].tags[].neckLine
object
Object containing neckline information
data.objects[].tags[].neckLine.id
string
Neckline unique ID
data.objects[].tags[].neckLine.name
string
Neckline description
data.objects[].tags[].neckLine.confidence
number
Neckline prediction confidence value (0~1)
Higher the confidence the closer value is to 1
data.objects[].tags[].fit
object
Fit prediction confidence value
data.objects[].tags[].fit.id
string
Fit unique ID
data.objects[].tags[].fit.name
string
Fit name
data.objects[].tags[].fit.confidence
number
Fit prediction confidence value (0~1)
Higher the confidence the closer the value is to 1
data.objects[].tags[].shape
object
Object containing shape information
data.objects[].tags[].shape.id
string
Shape unique ID
data.objects[].tags[].shape.name
string
Shape description
data.objects[].tags[].shape.confidence
number
Shape prediction confidence value (0~1) Higher the confidence the closer the value is to 1
data.objects[].tags[].heelHeight
object
Object containing heel height information for shoes
data.objects[].tags[].heelHeight.id
string
Heel height unique ID for shoes
data.objects[].tags[].heelHeight.name
string
Heel height description for shoes
data.objects[].tags[].heelHeight.confidence
number
Heel height prediction confidence value (0~1) for Higher the confidence the closer the value is to 1
data.objects[].tags[].toeType
object
Object containing toe type information for shoes
data.objects[].tags[].toeType.id
string
Toe type unique ID for shoes
data.objects[].tags[].toeType.name
string
Toe type description for shoes
data.objects[].tags[].toeType.confidence
number
Toe type prediction confidence value (0~1) for shoes Higher the confidence the closer the value is to 1
data.objects[].tags[].heelShape.id
string
Heel shape unique ID for shoes
data.objects[].tags[].heelShape.name
string
Heel shape description for shoes
data.objects[].tags[].heelShape.confidence
number
Heel shape prediction confidence value (0~1) for Higher the confidence the closer the value is to 1
data.objects[].tags[].soleType.id
string
Sole type unique ID for shoes
data.objects[].tags[].soleType.name
string
Sole type description for shoes
data.objects[].tags[].soleType.confidence
number
Sole type prediction confidence value (0~1) for shoes Higher the confidence the closer the value is to 1
data.notmatchedobjects[]
object array
Array of objects containing attribute details of items other than the ones specified in context.detection field, if the field is not empty.
error
object
Object containing error information if any
error.code
number
(If error has occurred) Error code
error.message
string
(If error has occurred) Error message
Response Example
If fashion items were detected from the input image
{
"data": {
"imageInfo": [
{
"id": "T0449",
"name": "nonFashion"
}
],
"genderInfo":
{
"id": "T0418",
"name": "male"
}
"objects": [
{
"type": "CLOTHING",
"tags": [
{
"category": {
"id": "T0360",
"name": "jacket"
},
"position": {
"x": 0.5033992828,
"y": 0.5039257706
},
"item": {
"id": "T0345",
"name": "balmacaan-jacket",
"confidence": 0.7787289023
},
"colors": [
{
"id": "T0326",
"name": "beige",
"confidence": 0.8287289023
}
],
"colorDetails": [
{
"code": "#e5ccaf",
"name": "gray-sand",
"pixelFraction": 1.0
}
],
"prints": [
{
"id": "T0198",
"name": "solid",
"confidence": 0.9973957539
}
],
"looks": [
{
"id": "T0141",
"name": "casual",
"confidence": 0.8063992262
}
],
"textures": [
{
"id": "T0220",
"name": "synthetic/polyester",
"confidence": 0.659104526
}
],
"details": [
{
"id": "T0097",
"name": "buttoned",
"confidence": 0.89336586
}
],
"length": {
"id": "T0468",
"name": "half",
"confidence": 0.5978361964
},
"sleeveLength": {
"id": "T0082",
"name": "long-sleeve",
"confidence": 0.9963954091
},
"neckLine": {
"id": "T0342",
"name": "hood",
"confidence": 0.6770937443
},
"fit": {
"id": "T0480",
"name": "skinny",
"confidence": 0.5594306588
},
"shape": {
"id": "T0493",
"name": "tapered",
"confidence": 0.9758504033
}
}
]
},
{
"type": "SHOES",
"tags": [
{
"category": {
"id": "T0495",
"name": "boots"
},
"position": {
"x": 0.3587198853,
"y": 0.8293977976
},
"item": {
"id": "T0503",
"name": "ankle-boots",
"confidence": 0.3421388268
},
"colors": [],
"heelHeight": {
"id": "T0536",
"name": "high(7~)",
"confidence": 0.609870851
},
"toeType": {
"id": "T0537",
"name": "round-toe",
"confidence": 0.5348223448
}
"heelShape": {
"id": "T0539",
"name": "stiletto",
"confidence": 0.7321234598
}
"soleType": {
"id": "T0542",
"name": "platform",
"confidence": 0.6345823217
}
}
]
}
]
},
"notMatchedObject": [
{
"type": "CLOTHING",
"tags": [
{
"category": {
"id": "T0360",
"name": "jacket"
},
"position": {
"x": 0.5033992828,
"y": 0.5039257706
},
"item": {
"id": "T0345",
"name": "balmacaan-jacket",
"confidence": 0.7787289023
},
"colors": [
{
"id": "T0326",
"name": "beige",
"confidence": 0.8287289023
}
],
"colorDetails": [
{
"code": "#e5ccaf",
"name": "gray-sand",
"pixelFraction": 1.0
}
],
"prints": [
{
"id": "T0198",
"name": "solid",
"confidence": 0.9973957539
}
],
"looks": [
{
"id": "T0141",
"name": "casual",
"confidence": 0.8063992262
}
],
"textures": [
{
"id": "T0220",
"name": "synthetic/polyester",
"confidence": 0.659104526
}
],
"details": [
{
"id": "T0097",
"name": "buttoned",
"confidence": 0.89336586
}
],
"length": {
"id": "T0468",
"name": "half",
"confidence": 0.5978361964
},
"sleeveLength": {
"id": "T0082",
"name": "long-sleeve",
"confidence": 0.9963954091
},
"neckLine": {
"id": "T0342",
"name": "hood",
"confidence": 0.6770937443
},
"fit": {
"id": "T0480",
"name": "skinny",
"confidence": 0.5594306588
},
"shape": {
"id": "T0493",
"name": "tapered",
"confidence": 0.9758504033
}
}
]
},
{
"type": "SHOES",
"tags": [
{
"category": {
"id": "T0495",
"name": "boots"
},
"position": {
"x": 0.3587198853,
"y": 0.8293977976
},
"item": {
"id": "T0503",
"name": "ankle-boots",
"confidence": 0.3421388268
},
"colors": [],
"heelHeight": {
"id": "T0536",
"name": "high(7~)",
"confidence": 0.609870851
},
"toeType": {
"id": "T0537",
"name": "round-toe",
"confidence": 0.5348223448
}
"heelShape": {
"id": "T0540",
"name": "ball",
"confidence": 0.6543234342
}
"soleType": {
"id": "T0543",
"name": "chunky-sneaker",
"confidence": 0.5677854423
}
}
]
}
]
},
},
"error": null,
"status": "ok"
}If no items were detected from the input image
{
"data": {
"imageInfo": [],
"objects": [
{
"type": "NODETC",
"tags": []
}
]
},
"error": null,
"status": "ok"
}In case of error occurrence
{
"data": {},
"status": "fail",
"error": {
"code": 803,
"message": "Fail to prove image from url."
}
}Last updated
Was this helpful?