Skip to content

ImagesApi.md

Latest commit

 

History

History
202 lines (147 loc) · 8.92 KB

ImagesApi.md

File metadata and controls

202 lines (147 loc) · 8.92 KB
Raw

MatroidPublicApi.ImagesApi

All URIs are relative to https://app.matroid.com

Method HTTP request Description
apiV1DetectorsDetectorIdClassifyImagePost POST /api/v1/detectors/{detectorId}/classify_image Classify image
apiV1DetectorsDetectorIdGetSaliencyMapPost POST /api/v1/detectors/{detectorId}/get_saliency_map Get saliency map
apiV1LocalizePost POST /api/v1/localize Localize objects in images

apiV1DetectorsDetectorIdClassifyImagePost

ApiV1DetectorsDetectorIdClassifyImagePost200Response apiV1DetectorsDetectorIdClassifyImagePost(detectorId, opts)

Classify image

Classifies one or more images using a detector. Note, the `classify_image` API call works for localizers as well and should be preferred over `localize_image`.

Example

import MatroidPublicApi from 'matroid_public_api';
let defaultClient = MatroidPublicApi.ApiClient.instance;
// Configure Bearer (JWT) access token for authorization: bearerAuth
let bearerAuth = defaultClient.authentications['bearerAuth'];
bearerAuth.accessToken = "YOUR ACCESS TOKEN"

let apiInstance = new MatroidPublicApi.ImagesApi();
let detectorId = "detectorId_example"; // String | ID of the detector to use for classification
let opts = {
  'url': ["null"], // [String] | URL(s) of image(s) to classify
  'file': ["null"], // [File] | Image file(s) to classify
  'localizationConfidence': 0.5, // Number | Minimum confidence threshold for localizations
  'numResults': 20, // Number | Maximum number of results to return
  'textLabels': ["null"] // [String] | Optional list of labels for open set detection
};
apiInstance.apiV1DetectorsDetectorIdClassifyImagePost(detectorId, opts, (error, data, response) => {
  if (error) {
    console.error(error);
  } else {
    console.log('API called successfully. Returned data: ' + data);
  }
});

Parameters

Name Type Description Notes
detectorId String ID of the detector to use for classification
url [String] URL(s) of image(s) to classify [optional]
file [File] Image file(s) to classify [optional]
localizationConfidence Number Minimum confidence threshold for localizations [optional] [default to 0.5]
numResults Number Maximum number of results to return [optional] [default to 20]
textLabels [String] Optional list of labels for open set detection [optional]

Return type

ApiV1DetectorsDetectorIdClassifyImagePost200Response

Authorization

bearerAuth

HTTP request headers

  • Content-Type: multipart/form-data
  • Accept: application/json

apiV1DetectorsDetectorIdGetSaliencyMapPost

ApiV1DetectorsDetectorIdGetSaliencyMapPost200Response apiV1DetectorsDetectorIdGetSaliencyMapPost(detectorId, version, classIdx, opts)

Get saliency map

Generate a saliency map for an image using a specified detector. One of file, imageId, or url must be provided.

Example

import MatroidPublicApi from 'matroid_public_api';
let defaultClient = MatroidPublicApi.ApiClient.instance;
// Configure Bearer (JWT) access token for authorization: bearerAuth
let bearerAuth = defaultClient.authentications['bearerAuth'];
bearerAuth.accessToken = "YOUR ACCESS TOKEN"

let apiInstance = new MatroidPublicApi.ImagesApi();
let detectorId = "detectorId_example"; // String | ID of the detector to use
let version = "version_example"; // String | API version identifier (e.g. \"v1\")
let classIdx = 3.4; // Number | Class index for saliency map generation
let opts = {
  'file': "/path/to/file", // File | Image file to process
  'imageId': "imageId_example", // String | ID of an existing image
  'url': "url_example" // String | URL of the image to process
};
apiInstance.apiV1DetectorsDetectorIdGetSaliencyMapPost(detectorId, version, classIdx, opts, (error, data, response) => {
  if (error) {
    console.error(error);
  } else {
    console.log('API called successfully. Returned data: ' + data);
  }
});

Parameters

Name Type Description Notes
detectorId String ID of the detector to use
version String API version identifier (e.g. "v1")
classIdx Number Class index for saliency map generation
file File Image file to process [optional]
imageId String ID of an existing image [optional]
url String URL of the image to process [optional]

Return type

ApiV1DetectorsDetectorIdGetSaliencyMapPost200Response

Authorization

bearerAuth

HTTP request headers

  • Content-Type: multipart/form-data
  • Accept: application/json

apiV1LocalizePost

ApiV1LocalizePost200Response apiV1LocalizePost(localizer, localizerLabel, opts)

Localize objects in images

This API is very similar to the classify endpoint. However, it can be used to update bounding boxes of existing training images by supplying `update=true`, `labelId`, and one of `imageId` or `imageIds`. It also has access to the internal face localizer (`localizer="DEFAULT_FACE"` and `localizerLabel="face"`). After receiving the results, perform the actual update using the Update Annotations endpoint. Note, the `classify_image` API call works for localizers as well and should be preferred over `localize_image` for inference.

Example

import MatroidPublicApi from 'matroid_public_api';
let defaultClient = MatroidPublicApi.ApiClient.instance;
// Configure Bearer (JWT) access token for authorization: bearerAuth
let bearerAuth = defaultClient.authentications['bearerAuth'];
bearerAuth.accessToken = "YOUR ACCESS TOKEN"

let apiInstance = new MatroidPublicApi.ImagesApi();
let localizer = "localizer_example"; // String | Unique ID of the detector (must support localization) or \\\"face\\\"
let localizerLabel = "localizerLabel_example"; // String | Name of the label to detect
let opts = {
  'confidence': 3.4, // Number | Minimum confidence level for predictions
  'file': "/path/to/file", // File | One or more image files
  'url': ["null"], // [String] | One or more image URLs
  'update': false, // Boolean | Whether to update the bounding boxes of existing training images.
  'maxFaces': 56, // Number | Maximum number of face localizations per image. Only relevant when using the default face localizer (i.e. localizer=\\\"DEFAULT_FACE\\\" and localizerLabel=\\\"face\\\")
  'labelId': "labelId_example", // String | Unique ID of the label for which to update images. Only relevant when updating (i.e. update=true)
  'imageId': "imageId_example", // String | Image ID. Only relevant when updating (i.e. update=true)
  'imageIds': ["null"] // [String] | One or more image IDs. Only relevant when updating (i.e. update=true)
};
apiInstance.apiV1LocalizePost(localizer, localizerLabel, opts, (error, data, response) => {
  if (error) {
    console.error(error);
  } else {
    console.log('API called successfully. Returned data: ' + data);
  }
});

Parameters

Name Type Description Notes
localizer String Unique ID of the detector (must support localization) or \"face\"
localizerLabel String Name of the label to detect
confidence Number Minimum confidence level for predictions [optional]
file File One or more image files [optional]
url [String] One or more image URLs [optional]
update Boolean Whether to update the bounding boxes of existing training images. [optional] [default to false]
maxFaces Number Maximum number of face localizations per image. Only relevant when using the default face localizer (i.e. localizer=\"DEFAULT_FACE\" and localizerLabel=\"face\") [optional]
labelId String Unique ID of the label for which to update images. Only relevant when updating (i.e. update=true) [optional]
imageId String Image ID. Only relevant when updating (i.e. update=true) [optional]
imageIds [String] One or more image IDs. Only relevant when updating (i.e. update=true) [optional]

Return type

ApiV1LocalizePost200Response

Authorization

bearerAuth

HTTP request headers

  • Content-Type: multipart/form-data
  • Accept: application/json