gradio_mediagallery

Media Gallery Explorer with Metadata Image Extraction for Gradio UI

MediaGallery with Metadata Extraction

To Test:

  1. Use the FolderExplorer on the left to select a folder containing images with metadata.
  2. Click on an image in the Media Gallery to open the preview mode.
  3. In the preview toolbar, click the 'Info' icon (ⓘ) to open the metadata popup.
  4. Click the 'Load Metadata' button inside the popup.
  5. The fields in the Metadata Viewer below will be populated with the data from the image.
{}
[]

Metadata Viewer

0 1

Installation

pip install gradio_mediagallery

Usage

from typing import Any, List
import gradio as gr
from gradio_folderexplorer import FolderExplorer
from gradio_folderexplorer.helpers import load_media_from_folder
from gradio_mediagallery import MediaGallery
from gradio_mediagallery.helpers import transfer_metadata

# Configuration constant for the root directory containing media files
ROOT_DIR_PATH = "./src/examples"

def handle_load_metadata(image_data: gr.EventData) -> List[Any]:
    """
    Processes image metadata by calling the `transfer_metadata` helper.

    Args:
        image_data (gr.EventData): Event data containing metadata from the MediaGallery component.

    Returns:
        List[Any]: A list of values to populate the output fields, or skipped updates if no data is provided.
    """
    if not image_data or not hasattr(image_data, "_data"):
        return [gr.skip()] * len(output_fields)

    return transfer_metadata(
        output_fields=output_fields,
        metadata=image_data._data,
        remove_prefix_from_keys=True
    )

# UI layout and logic
with gr.Blocks(theme=gr.themes.Ocean()) as demo:
    """
    A Gradio interface for browsing and displaying media files with metadata extraction.
    """
    gr.Markdown("# MediaGallery with Metadata Extraction")
    gr.Markdown(
        """
        **To Test:**
        1. Use the **FolderExplorer** on the left to select a folder containing images with metadata.
        2. Click on an image in the **Media Gallery** to open the preview mode.
        3. In the preview toolbar, click the 'Info' icon (ⓘ) to open the metadata popup.
        4. Click the **'Load Metadata'** button inside the popup.
        5. The fields in the **Metadata Viewer** below will be populated with the data from the image.
        """
    )
    with gr.Row(equal_height=True):
        with gr.Column(scale=1, min_width=300):
            folder_explorer = FolderExplorer(
                label="Select a Folder",
                root_dir=ROOT_DIR_PATH,
                value=ROOT_DIR_PATH
            )

        with gr.Column(scale=3):
            gallery = MediaGallery(
                label="Media in Folder",
                columns=6,
                height="auto",
                preview=False,
                show_download_button=False,
                only_custom_metadata=False,
                popup_metadata_width="40%",
            )

    gr.Markdown("## Metadata Viewer")
    with gr.Row():
        model_box = gr.Textbox(label="Model")
        fnumber_box = gr.Textbox(label="FNumber")
        iso_box = gr.Textbox(label="ISOSpeedRatings")
        s_churn = gr.Slider(label="Schurn", minimum=0.0, maximum=1.0, step=0.01)
        description_box = gr.Textbox(label="Description", lines=2)

    # Event handling
    output_fields = [
        model_box,
        fnumber_box,
        iso_box,
        s_churn,
        description_box
    ]

    # Populate the gallery when the folder changes
    folder_explorer.change(
        fn=load_media_from_folder,
        inputs=folder_explorer,
        outputs=gallery
    )

    # Populate the gallery on initial load
    demo.load(
        fn=load_media_from_folder,
        inputs=folder_explorer,
        outputs=gallery
    )

    # Handle the load_metadata event from MediaGallery
    gallery.load_metadata(
        fn=handle_load_metadata,
        inputs=None,
        outputs=output_fields
    )

if __name__ == "__main__":
    """
    Launches the Gradio interface in debug mode.
    """
    demo.launch(debug=True)

MediaGallery

Initialization

Parameters
value: Sequence[
        np.ndarray | PIL.Image.Image | str | Path | tuple
    ]
    | Callable
    | None
default = None

Initial list of images or videos, or a function to generate them.

file_types: list[str] | None
default = None

List of allowed file extensions or types for uploads (e.g., ['image', '.mp4']).

label: str | I18nData | None
default = None

Label displayed above the component.

every: Timer | float | None
default = None

Interval or Timer to refresh `value` if it's a function.

inputs: Component | Sequence[Component] | set[Component] | None
default = None

Components used as inputs to recalculate `value` if it's a function.

show_label: bool | None
default = None

Whether to display the label.

container: bool
default = True

Whether to place the component in a padded container.

scale: int | None
default = None

Relative size compared to adjacent components.

min_width: int
default = 160

Minimum pixel width of the component.

visible: bool | Literal["hidden"]
default = True

Whether the component is visible or hidden.

elem_id: str | None
default = None

HTML ID for the component.

elem_classes: list[str] | str | None
default = None

HTML classes for the component.

render: bool
default = True

Whether to render the component in the Blocks context.

key: int | str | tuple[int | str, ...] | None
default = None

Identifier for preserving component state across re-renders.

preserved_by_key: list[str] | str | None
default = "value"

Parameters to preserve during re-renders.

columns: int | None
default = 2

Number of columns in the grid.

rows: int | None
default = None

Number of rows in the grid.

height: int | float | str | None
default = None

Height of the gallery in pixels or CSS units.

allow_preview: bool
default = True

Whether images can be enlarged on click.

preview: bool | None
default = None

Whether to start in preview mode (requires allow_preview=True).

selected_index: int | None
default = None

Index of the initially selected media item.

object_fit: Literal[
        "contain", "cover", "fill", "none", "scale-down"
    ]
    | None
default = None

CSS object-fit for thumbnails ("contain", "cover", etc.).

show_share_button: bool | None
default = None

Whether to show a share button (auto-enabled on Hugging Face Spaces).

show_download_button: bool | None
default = True

Whether to show a download button for the selected media.

interactive: bool | None
default = None

Whether the gallery allows uploads.

type: Literal["numpy", "pil", "filepath"]
default = "filepath"

Format for images passed to the prediction function ("numpy", "pil", "filepath").

show_fullscreen_button: bool
default = True

Whether to show a fullscreen button.

only_custom_metadata: bool
default = True

Whether to filter out technical EXIF metadata in the popup.

popup_metadata_width: int | str
default = 500

Width of the metadata popup (pixels or CSS string).

Events

Parameters
select

Event listener for when the user selects or deselects the MediaGallery. Uses event data gradio.SelectData to carry `value` referring to the label of the MediaGallery, and `selected` to refer to state of the MediaGallery. See EventData documentation on how to use this event data

change

Triggered when the value of the MediaGallery changes either because of user input (e.g. a user types in a textbox) OR because of a function update (e.g. an image receives a value from the output of an event trigger). See `.input()` for a listener that is only triggered by user input.

delete

This listener is triggered when the user deletes and item from the MediaGallery. Uses event data gradio.DeletedFileData to carry `value` referring to the file that was deleted as an instance of FileData. See EventData documentation on how to use this event data

preview_close

Triggered when the MediaGallery preview is closed by the user.

preview_open

Triggered when the MediaGallery preview is opened by the user.

load_metadata

Triggered when the user clicks the 'Load Metadata' button in the metadata popup. Returns a dictionary of image metadata.

User function

The impact on the users predict function varies depending on whether the component is used as an input or output for an event (or both).

  • When used as an Input, the component only impacts the input signature of the user function.
  • When used as an output, the component only impacts the return signature of the user function.

The code snippet below is accurate in cases where the component is used as both an input and an output.

  • As input: Is passed, list of tuples containing file paths and captions, or None if payload is None.
  • As output: Should return, list of media items (images, videos, or tuples with captions).
def predict(
   value: Any
) -> list | None:
   return value