The similar image clustering feature groups images that you have indexed in a dataset into clusters based on visual similarity. This feature is useful for scenarios such as deduplicating images or selecting the best shots. For example, you can use it to filter burst photos in an album.
Operation description
-
Before calling this operation, review the billing methods and pricing of Intelligent Media Management (IMM).
Important The execution time of asynchronous tasks is not guaranteed. -
Before calling this operation, index files to a dataset. You can index files by attaching a data source using CreateBinding, or by actively indexing files using IndexFileMeta or BatchIndexFileMeta.
-
Each call to this operation incrementally processes the files in the specified
Dataset. You can call this operation periodically to process new files. -
After clustering completes, call the QuerySimilarImageClusters operation to retrieve the clustering results.
-
Each similar image cluster must contain at least two images. Deleting a file from a dataset changes the similar image clusters. If deleting an image reduces a cluster to fewer than two images, the cluster is automatically deleted.
-
This operation is asynchronous. After a task starts, its information is retained for only seven days. You cannot query the information after this period. Call the GetTask or ListTasks operation and use the returned
TaskIdto view task information. You can also set the Notification parameter to receive asynchronous notification messages about the task.
Try it now
Test
RAM authorization
|
Action |
Access level |
Resource type |
Condition key |
Dependent action |
|
imm:CreateSimilarImageClusteringTask |
create |
*Dataset
|
None | None |
Request parameters
|
Parameter |
Type |
Required |
Description |
Example |
| ProjectName |
string |
Yes |
The name of the project. For more information, see Create a project. |
test-project |
| DatasetName |
string |
Yes |
The name of the dataset. For more information, see Create a dataset. |
test-dataset |
| UserData |
string |
No |
The custom information. This information is returned in the asynchronous notification message to help you associate the message with your system. The value can be up to 2,048 bytes in length. |
test-data |
| Tags |
object |
No |
The custom tags that are used to search for and filter asynchronous tasks. |
{ "User": "Jane" } |
| Notification | Notification |
No |
The configuration of the notification message. For more information, click Notification. For more information about the format of asynchronous notification messages, see Asynchronous notification messages. |
Response elements
|
Element |
Type |
Description |
Example |
|
object |
The information about the similar image clustering task. |
||
| RequestId |
string |
The request ID. |
1B3D5E0A-D8B8-4DA0-8127-ED32C851**** |
| TaskId |
string |
The task ID. |
SimilarImageClustering-48d0a0f3-8459-47f4-b8af-ff49c64**** |
| EventId |
string |
The event ID. |
3BF-1UhtFyrua71eOkFlqYq23Co**** |
Examples
Success response
JSON format
{
"RequestId": "1B3D5E0A-D8B8-4DA0-8127-ED32C851****",
"TaskId": "SimilarImageClustering-48d0a0f3-8459-47f4-b8af-ff49c64****",
"EventId": "3BF-1UhtFyrua71eOkFlqYq23Co****"
}
Error codes
See Error Codes for a complete list.
Release notes
See Release Notes for a complete list.