ObjectDetection

Create a new instance

The job failed to process.

Create GenericImage instances for the images of this job.

Clean up temporary data produced by this job.

Dispatch the job to notify the BIIGLE instance of a failure.

Create the temporary directory for this request.

Creates the specified directory if it does not exist.

Get the path to the temporary directory.

Execute a Python command and return the path to a file containing the output.

Parse the output JSON files to get the array of annotations for each image.

Parse the output JSON file of a single image.

Create MAIA annotations from the training proposals.

Create an insert array for a MAIA annotation.

Apply the limit for the maximum number of annotations.

Insert one chunk of the MAIA annotations that should be created into the database.

Update the state of the MAIA job after processing the response.

Execute the job

Determine whether knowledge transfer should be performed in this job.

Bundle the training proposals to be sent to the GPU server.

Generate the training dataset for the object detection model.

Create the JSON file that is the input to the dataset generation script.

Perform training of object detection model.

Create the JSON file that is the input to the training script.

Perform inference with the trained object detection model.

Create the JSON file that is the input to the inference script.

Build the map from image ID to path of the cached image file.

Dispatch the job to store the object detection results.

Create GenericImage instances for the images of the knowledge transfer volume.