Package ai.djl.modality.cv

Class VisionLanguageInput

java.lang.Object

ai.djl.modality.cv.VisionLanguageInput

public class VisionLanguageInput
extends Object

The input container for a vision language model.

Constructor Summary

Constructors

Constructor	Description
VisionLanguageInput(Image image, String text)	Constructs a new ImageTextInput instance.
VisionLanguageInput(Image image, String[] candidates)	Constructs a new ImageTextInput instance.
VisionLanguageInput(Image image, String text, String[] candidates, String hypothesisTemplate)	Constructs a new ImageTextInput instance.

Method Summary

Modifier and Type	Method	Description
String[]	getCandidates()	Returns the candidate labels.
String	getHypothesisTemplate()	Returns the hypothesis template.
Image	getImage()	Returns the image input.
String	getText()	Returns the prompt text.
static VisionLanguageInput	parseInput(Input input)	Returns the ImageTextInput from the Input.
void	setCandidates(String[] candidates)	Sets the candidate labels.
void	setHypothesisTemplate(String hypothesisTemplate)	Sets the hypothesis template.
void	setImage(Image image)	Sets the image input.
void	setText(String text)	Sets the prompt text.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

VisionLanguageInput

public VisionLanguageInput(Image image,
 String text)

Constructs a new ImageTextInput instance.

Parameters:

image - the image input

text - the prompt

VisionLanguageInput

public VisionLanguageInput(Image image,
 String[] candidates)

Constructs a new ImageTextInput instance.

Parameters:

image - the image input

candidates - the candidate labels

VisionLanguageInput

public VisionLanguageInput(Image image,
 String text,
 String[] candidates,
 String hypothesisTemplate)

Constructs a new ImageTextInput instance.

Parameters:

image - the image input

text - the prompt

candidates - the candidate labels

hypothesisTemplate - the hypothesis template

Method Details

parseInput

public static VisionLanguageInput parseInput(Input input)
                                      throws TranslateException,
IOException

Returns the ImageTextInput from the Input.

Parameters:

input - the input object

Returns:

the ImageTextInput from the Input

Throws:

TranslateException - if the input is invalid

IOException - if failed to load image

getImage

public Image getImage()

Returns the image input.

Returns:

the image input

setImage

public void setImage(Image image)

Sets the image input.

Parameters:

image - the image input

getText

public String getText()

Returns the prompt text.

Returns:

the prompt text

setText

public void setText(String text)

Sets the prompt text.

Parameters:

text - the prompt text

getCandidates

public String[] getCandidates()

Returns the candidate labels.

Returns:

the candidate labels

setCandidates

public void setCandidates(String[] candidates)

Sets the candidate labels.

Parameters:

candidates - the candidate labels

getHypothesisTemplate

public String getHypothesisTemplate()

Returns the hypothesis template.

Returns:

the hypothesis template

setHypothesisTemplate

public void setHypothesisTemplate(String hypothesisTemplate)

Sets the hypothesis template.

Parameters:

hypothesisTemplate - the hypothesis template