🔶Generate Labeled Dataset

Function Description

This operator is used to convert annotated data in a specific format to COCO dataset format, which can be directly uploaded to the transfer CVAT annotation platform for training deep learning models.

Usage Scenarios

Prepare deep learning data, which can be further used to optimize annotation and manage model data.

Input Output

Input

Save Path： Specifies the folder where the original annotation data is located. The operator will look for .png files in this folder and expects each .png file to have a .json file with the same name (e.g., image1.png corresponds to image1.json).

  • If this input port is connected with data (i.e., a path is provided), this path will override the "Save Path" parameter set in the parameter panel and serve as the read path for original data.

  • If this input port is not connected with data (empty), the operator will use the path specified by "Save Path" in the parameter panel as the read path for original data.

Output

File Name： The main name of the actually saved COCO dataset file.

Parameter Description

The annotation name parameter must provide an array in the correct format, with each custom category name enclosed in English double quotes ("), such as ["Category A", "Category B"], otherwise it will prompt "Format failed".

Save Path

Parameter Description

Used to specify the target folder path for file saving.

Parameter Adjustment

  • Specify Path: Directly input a valid folder path, or click the icon on the right to select a folder path.

  • No Path Specified: Leave the path empty, and the operator will default to creating a file_save_dir subfolder in the .transfertech atom folder as the storage path. This is convenient for temporary saving or when you don’t care about the specific storage location.

Path Check: The operator will check whether the specified path is valid. If the path points to an existing file (rather than a folder) or the path is invalid, the operator will report an error. If the specified path doesn’t exist, it will attempt to create it.

Save Name

Parameter Description

Specify the save file name, no need to include file extension.

Parameter Adjustment

It’s recommended to use descriptive names for easy subsequent retrieval and management, avoiding special characters.

Default Name: data

Save Single File Only

Parameter Description

Sets the file saving strategy, save as single file or multiple files.

Parameter Adjustment

  • Disable (Default): Generate multiple files, automatically add timestamps after the specified filename to avoid duplicate names, suitable for scenarios that need to record historical data or multiple run results.

  • Enable: The operator will attempt to use the specified name as the filename to save data each time it runs. If the file already exists, new data will overwrite the old file. Suitable for scenarios that need fixed filename output, such as overwriting previous results.

Maximum Storage Memory

Parameter Description

Sets the maximum total size allowed for files stored in the save directory. When the total size of related files in the folder exceeds this limit, the software will attempt to delete the earliest created files of the same type in that directory to free up space.

Parameter Adjustment

Please set this value based on available disk space and data retention requirements:

  • Smaller Value： Will delete old files more frequently, saving disk space, but may cause historical data to be lost more quickly.

  • Larger Value： Allows saving more historical data, but will occupy more disk space.

Parameter Range

[0.00001, 10], Default: 10, Unit: GB

Annotation Names

Parameter Description

A string list used to define all possible object category names in the dataset. Each string in the list represents a category.

Parameter Adjustment

  • Custom Categories： Set this list according to requirements with custom category names. For example, if your project detects "cat", "dog", "person", you can set it to ["cat", "dog", "person"].

  • Category ID Mapping： The operator will automatically map the first name in the list to ID 1, the second to ID 2, and so on. However, in annotation files, the label field will correspond to category names using numerical strings starting from 0.

  • Default Value : If this list is empty, the operator will default to using ["0", "1", "2", "3", "4"] as category names, supporting up to 5 categories.