_Clarify_ai, _Convert_ai, _Denoise_ai, _Enhance_ai, _Explorer_ai, _Restore_ai, _Segment_ai, _SegmentObjects_ai, _Train_Convert_ai, _Train_Enhance_ai, _Train_Segment_ai, _Train_SegmentObjects_ai, Convert_ai, CreateTrainedFileDBRecord, Denoise_ai, Enhance_ai, Restore_ai, RunDistributedItem, RunDistributedItems, RunQueuedItem, RunQueuedItems, Segment_ai, SegmentObjects_ai, StartMxExecutor, Train_Convert_ai, Train_Enhance_ai, Train_Segment_ai, Train_SegmentObjects_ai

Convert_ai

Interactive command to this function: NIS.ai > Convert.ai

Description 

 Convert_ai(
   char *SourceChannelIndices,
   char *NetworkFileName
);

Runs the Convert.ai neural network using the defined parameters.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

char *NetworkFileName

Path to an existing neural network file.

See Also

• _Train_Convert_ai

CreateTrainedFileDBRecord

Description 

 CreateTrainedFileDBRecord(
   char *Filename
);

Writes the trained file from AI into the database.

Parameters

char *Filename

The path and the name of the file.

Denoise_ai

Interactive command to this function: NIS.ai > Denoise.ai

Description 

 Denoise_ai(
   char *SourceChannelIndices,
   int  NewDocument
);

Performs image denoising with the use of neural networks.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

int NewDocument

Creates a new image document once the process is finished.

0	The current image is overwritten.
1	A new image is created.

See Also

• _Denoise_ai

Enhance_ai

Interactive command to this function: NIS.ai > Enhance.ai

Description 

 Enhance_ai(
   char *SourceChannelIndices,
   char *NetworkFileName
);

Runs the Enhance.ai neural network using the defined parameters.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

char *NetworkFileName

Path to an existing neural network file.

See Also

• _Train_Enhance_ai

Restore_ai

Interactive command to this function: NIS.ai > Restore.ai

Description 

 Restore_ai(
   char *SourceChannelIndices,
   int  NewDocument
);

Runs the Restore.ai using the defined parameters.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

int NewDocument

Creates a new image document once the process is finished.

0	The current image is overwritten.
1	A new image is created.

See Also

• _Restore_ai

RunDistributedItem

Description 

 RunDistributedItem(
   char *Title
);

Sends the specified unprocessed training that was intended for distributed processing to the distributed server. For example when no distributed server was available at the time of the training definition.

Parameters

char *Title

Title name.

RunDistributedItems

Description 

 RunDistributedItems();

Sends all unprocessed training to the distributed server which was intended for the distribution processing. For example when no distributed server was available at the time of the training definition.

RunQueuedItem

Description 

 RunQueuedItem(
   char *Title
);

Runs Train AI on a specified queued AI definition.

Parameters

char *Title

Title name.

RunQueuedItems

Description 

 RunQueuedItems();

Runs Train AI on all queued AI definitions.

Segment_ai

Interactive command to this function: NIS.ai > Segment.ai

Description 

 Segment_ai(
   char *SourceChannelIndices,
   char *NetworkFileName,
   int  AdvMultiSmooth,
   double  AdvSmoothSize,
   double  AdvCleanSize,
   int  AdvFillHoles,
   double  AdvSeparateSize,
   int  AdvFilterSize,
   double  AdvFilterSizeLow,
   double  AdvFilterSizeHigh,
   int  AdvFilterCircularity,
   double  AdvFilterCircularityLow,
   double  AdvFilterCircularityHigh
);

Runs the Segment.ai neural network using the defined parameters.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

char *NetworkFileName

Path to an existing neural network file.

int AdvMultiSmooth

Sets the smooth processing.

double AdvSmoothSize

Smooth processing size [px].

double AdvCleanSize

Clean processing size [px].

int AdvFillHoles

Sets the fill holes processing.

double AdvSeparateSize

Separate processing size [px].

int AdvFilterSize

Sets the size filter.

double AdvFilterSizeLow

Low size filter value [px].

double AdvFilterSizeHigh

High size filter value [px].

int AdvFilterCircularity

Sets the circularity filter.

double AdvFilterCircularityLow

Low circularity filter value.

double AdvFilterCircularityHigh

High circularity filter value.

See Also

• _Train_Segment_ai

SegmentObjects_ai

Interactive command to this function: NIS.ai > Segment Objects.ai

Description 

 SegmentObjects_ai(
   char *SourceChannelIndices,
   char *NetworkFileName,
   int  AdvSmoothObjectsRadius,
   double  AdvCleanSize,
   int  AdvFillHoles,
   int  AdvFilterSize,
   double  AdvFilterSizeLow,
   double  AdvFilterSizeHigh,
   int  AdvFilterCircularity,
   double  AdvFilterCircularityLow,
   double  AdvFilterCircularityHigh
);

Runs the Segment.ai neural network using the defined parameters.

Parameters

char *SourceChannelIndices

Indices of the source channels used for neural training. If no channel is selected, then no channel is processed.

char *NetworkFileName

Path to an existing neural network file.

int AdvSmoothObjectsRadius

Sets the advanced objects smooth radius.

double AdvCleanSize

Clean processing size [px].

int AdvFillHoles

Sets the fill holes processing.

int AdvFilterSize

Sets the size filter.

double AdvFilterSizeLow

Low size filter value [px].

double AdvFilterSizeHigh

High size filter value [px].

int AdvFilterCircularity

Sets the circularity filter.

double AdvFilterCircularityLow

Low circularity filter value.

double AdvFilterCircularityHigh

High circularity filter value.

See Also

• _Train_Segment_ai

StartMxExecutor

Description 

 StartMxExecutor();

Starts the MX executor.

Train_Convert_ai

Interactive command to this function: NIS.ai > Train Convert.ai

Description 

 Train_Convert_ai(
   char *DatasetFileNames,
   int  SourceChannelNames,
   int  GroundTruthChannelNames,
   char *NetworkOptions,
   char *NetworkInputFileName,
   char *NetworkOutputFileName,
   char *NetworkName,
   char *GpuPciBusIds
);

Starts training of the Convert.ai neural network on the current dataset.

Parameters

char *DatasetFileNames

Path(s) to the training data file(s).

int SourceChannelNames

Indices of the source channels used for neural training.

int GroundTruthChannelNames

Ground truth channel channel used for neural training.

char *NetworkOptions

More neural training options.

closeOnFinish	Closes the window after the network training is finished.
test	Set to a value from 0 to 100. The value specifies what percentage of the training dataset will be used for testing the network. This data will only be used for testing of the performance of the network, but not for training. Do not use this parameter on datasets which contains only a few frames of data so that you don't unnecessarily reduce the amount of the training data.
flips	Flips images during the neural training.
epochs	Number of neural training iteration.
gamma	Set to value from range <1.0, 5.0>. Changes image gamma randomly from range <1.0/value, value>. When set to 1.0, no augmentation is performed.
screenshot	Saves a screenshot of the training graph.

char *NetworkInputFileName

Path to an existing neural network file.

char *NetworkOutputFileName

Path where the neural network file will be saved.

char *NetworkName

Name of the neural network used in the database.

char *GpuPciBusIds

Sets the GPU used for training. It is a list of comma separated numbers in this format XX:XX.X where X is any digit. The digits do not need to be the same, e.g. 21:00.0. If the list is empty, all available GPUs are used.

See Also

• _Train_Convert_ai

Train_Enhance_ai

Interactive command to this function: NIS.ai > Train Enhance.ai

Description 

 Train_Enhance_ai(
   char *DatasetFileNames,
   int  SourceChannelNames,
   int  GroundTruthChannelNames,
   char *NetworkOptions,
   char *NetworkInputFileName,
   char *NetworkOutputFileName,
   char *NetworkName,
   char *GpuPciBusIds
);

Starts training of the Enhance.ai neural network on the current dataset.

Parameters

char *DatasetFileNames

Path(s) to the training data file(s).

int SourceChannelNames

Indices of the source channels used for neural training.

int GroundTruthChannelNames

Ground truth channel channel used for neural training.

char *NetworkOptions

More neural training options.

closeOnFinish	Closes the window after the network training is finished.
test	Set to a value from 0 to 100. The value specifies what percentage of the training dataset will be used for testing the network. This data will only be used for testing of the performance of the network, but not for training. Do not use this parameter on datasets which contains only a few frames of data so that you don't unnecessarily reduce the amount of the training data.
flips	Flips images during the neural training.
epochs	Number of neural training iteration.
gamma	Set to value from range <1.0, 5.0>. Changes image gamma randomly from range <1.0/value, value>. When set to 1.0, no augmentation is performed.
screenshot	Saves a screenshot of the training graph.

char *NetworkInputFileName

Path to an existing neural network file.

char *NetworkOutputFileName

Path where the neural network file will be saved.

char *NetworkName

Name of the neural network used in the database.

char *GpuPciBusIds

Sets the GPU used for training. It is a list of comma separated numbers in this format XX:XX.X where X is any digit. The digits do not need to be the same, e.g. 21:00.0. If the list is empty, all available GPUs are used.

See Also

• _Train_Enhance_ai

Train_Segment_ai

Interactive command to this function: NIS.ai > Train Segment.ai

Description 

 Train_Segment_ai(
   char *DatasetFileNames,
   int  SourceChannelNames,
   char *GroundTruthBinaryNames,
   char *NetworkOptions,
   char *NetworkInputFileName,
   char *NetworkOutputFileName,
   char *NetworkName,
   char *GpuPciBusIds
);

Starts training of the Segment.ai neural network on the current dataset.

Parameters

char *DatasetFileNames

Path(s) to the training data file(s).

int SourceChannelNames

Indices of the source channels used for neural training.

char *GroundTruthBinaryNames

Path to the binary layers used for neural training.

char *NetworkOptions

More neural training options.

closeOnFinish	Closes the window after the network training is finished.
test	Set to a value from 0 to 100. The value specifies what percentage of the training dataset will be used for testing the network. This data will only be used for testing of the performance of the network, but not for training. Do not use this parameter on datasets which contains only a few frames of data so that you don't unnecessarily reduce the amount of the training data.
flips	Flips images during the neural training.
epochs	Number of neural training iteration.
gamma	Set to value from range <1.0, 5.0>. Changes image gamma randomly from range <1.0/value, value>. When set to 1.0, no augmentation is performed.
screenshot	Saves a screenshot of the training graph.

char *NetworkInputFileName

Path to an existing neural network file.

char *NetworkOutputFileName

Path where the neural network file will be saved.

char *NetworkName

Name of the neural network used in the database.

char *GpuPciBusIds

Sets the GPU used for training. It is a list of comma separated numbers in this format XX:XX.X where X is any digit. The digits do not need to be the same, e.g. 21:00.0. If the list is empty, all available GPUs are used.

See Also

• _Train_Segment_ai

Train_SegmentObjects_ai

Interactive command to this function: NIS.ai > Train Segment Objects.ai

Description 

 Train_SegmentObjects_ai(
   char *DatasetFileNames,
   int  SourceChannelNames,
   char *GroundTruthBinaryNames,
   char *NetworkOptions,
   char *NetworkInputFileName,
   char *NetworkOutputFileName,
   char *NetworkName,
   char *GpuPciBusIds
);

Starts training of the Segment Objects.ai neural network on the current dataset.

Parameters

char *DatasetFileNames

Path(s) to the training data file(s).

int SourceChannelNames

Indices of the source channels used for neural training.

char *GroundTruthBinaryNames

Path to the binary layers used for neural training.

char *NetworkOptions

More neural training options.

closeOnFinish	Closes the window after the network training is finished.
test	Set to a value from 0 to 100. The value specifies what percentage of the training dataset will be used for testing the network. This data will only be used for testing of the performance of the network, but not for training. Do not use this parameter on datasets which contains only a few frames of data so that you don't unnecessarily reduce the amount of the training data.
flips	Flips images during the neural training.
epochs	Number of neural training iteration.
gamma	Set to value from range <1.0, 5.0>. Changes image gamma randomly from range <1.0/value, value>. When set to 1.0, no augmentation is performed.
screenshot	Saves a screenshot of the training graph.

char *NetworkInputFileName

Path to an existing neural network file.

char *NetworkOutputFileName

Path where the neural network file will be saved.

char *NetworkName

Name of the neural network used in the database.

char *GpuPciBusIds

Sets the GPU used for training. It is a list of comma separated numbers in this format XX:XX.X where X is any digit. The digits do not need to be the same, e.g. 21:00.0. If the list is empty, all available GPUs are used.

See Also

• _Train_SegmentObjects_ai

_Clarify_ai

This function runs the NIS.ai > Clarify.ai command.

Description 

 _Clarify_ai();

Opens the Clarify.ai dialog window.

_Convert_ai

This function runs the NIS.ai > Convert.ai command.

Description 

 _Convert_ai();

Opens a window for running a trained neural network (*.cai) on the current dataset.

_Denoise_ai

This function runs the NIS.ai > Denoise.ai command.

Description 

 _Denoise_ai();

Opens the Denoise.ai dialog window suitable for image denoising with the use of neural networks.

See Also

• Denoise_ai

_Enhance_ai

This function runs the NIS.ai > Enhance.ai command.

Description 

 _Enhance_ai();

Opens a window for running a trained neural network (*.eai) on the current dataset.

_Explorer_ai

This function runs the NIS.ai > Explorer.ai command.

Description 

 _Explorer_ai();

Opens the NIS.ai Explorer panel.

_Restore_ai

This function runs the NIS.ai > Restore.ai command.

Description 

 _Restore_ai();

Opens the Restore.ai dialog window.

See Also

• Restore_ai

_Segment_ai

This function runs the NIS.ai > Segment.ai command.

Description 

 _Segment_ai();

Opens a window for running a trained neural network (*.sai) on the current dataset.

_SegmentObjects_ai

This function runs the NIS.ai > Segment Objects.ai command.

Description 

 _SegmentObjects_ai();

Opens a window for running a trained neural network (*.oai) on the current dataset.

_Train_Convert_ai

This function runs the NIS.ai > Train Convert.ai command.

Description 

 _Train_Convert_ai();

Enables the user to train a neural network on the current dataset.

See Also

• Train_Convert_ai

_Train_Enhance_ai

This function runs the NIS.ai > Train Enhance.ai command.

Description 

 _Train_Enhance_ai();

Enables the user to train a neural network on the current dataset.

See Also

• Train_Enhance_ai

_Train_Segment_ai

This function runs the NIS.ai > Train Segment.ai command.

Description 

 _Train_Segment_ai();

Enables the user to train a neural network on the current dataset.

See Also

• Train_Segment_ai

_Train_SegmentObjects_ai

This function runs the NIS.ai > Train Segment Objects.ai command.

Description 

 _Train_SegmentObjects_ai();

Enables the user to train a neural network on the current dataset.

See Also

• Train_SegmentObjects_ai