Home

The Home view is where you browse inspection images captured from your production line. You can filter images by time, camera, annotation status, predictions, and more. From here you start annotation workflows, tag images, or select images for model comparison.

Layout

Home is divided into two areas:

• Left sidebar — filters, actions, and tag management
• Main area — image grid with pagination and selection controls

Filter pane

The left sidebar contains the filter pane. Filters narrow which images appear in the grid.

Tip

After changing filters, click Apply filters to update the image grid. Filters are not applied automatically unless Auto-update is enabled.

Time range

Filter	Description
From	Start of the time range. Defaults to midnight today (or yesterday if before 2 AM).
To	End of the time range. Leave empty to include the most recent images.

Image identity

Filter	Description
Rig	Filter by rig ID — the physical inspection station.
Cam	Filter by camera ID within a rig.
Img	Filter by image ID — the position within a capture sequence.
Frame trigger	Filter by frame trigger ID — links all images from a single capture event.

Note

All text filters support negation. Prefix a value with ! to exclude it. For example, entering !3 in the Camera filter excludes camera 3.

Predictions and annotations

Filter	Description
Predictions	Filter by prediction values. Format: label:value. Use !label:value to exclude.
Annotations	Filter by annotation values. Same format as predictions.
Confidence min / max	Show only images with prediction confidence within this percentage range (0–100).
Model	Restrict results to a specific AI model, or select “All”.
Annotated by	Show only images annotated by a specific user.

Display options

Filter	Description
Show / hide annotated	Choose to show all images, only annotated, or only unannotated.
Show / hide predicted	Choose to show all images, only predicted, or only unpredicted.
Only show prediction above threshold	A slider (0–100%) that controls which predictions are visible on image cards. Predictions below this confidence are hidden from the display.
Hide skipped annotations	When enabled, images that have been skipped during annotation are hidden. On by default.
Auto-update	Polls for new images every second. Useful when monitoring a live production line.

Caution

Auto-update queries the database every second. Disable it when you are not actively monitoring incoming images.

Tags and additional fields

Filter	Description
Tag	Filter by custom tags applied to images.
Additional data fields	If configured, extra fields from the capture pipeline appear as filters (for example, a barcode or lot number).

Filter actions

Button	What it does
Apply filters	Applies the current filter settings and refreshes the image grid.
Clear filters	Resets all filters to their defaults.

Tip

Filters are saved in your browser. When you return to EyVz, your previous filter settings are restored.

Image grid

The main area displays images as a grid of cards, five columns wide. Above the grid you control pagination and selection.

Pagination

Control	Description
Images per page	Set how many images to show per page. Default: 10.
Page navigation	Jump to first, previous, next, or last page, or type a specific page number.
Displaying X–Y of Z	Shows which images are currently visible out of the total result count.

Selection controls

Button	What it does
Select displayed	Selects all images on the current page.
Select all	Selects all images matching the current filters (up to 10,000). If more exist, a confirmation dialog appears.
Clear page	Deselects images on the current page only.
Clear all selected	Deselects everything.

A counter shows how many images are currently selected.

Image cards

Each card in the grid shows:

Element	Description
Image thumbnail	The captured image. Click to open it in the annotation view (read-only).
Sticker number	A barcode value identifying the physical item, shown when available.
ID badges	Rig ID, Camera ID, Image ID, Trigger ID, and capture date.
Annotation status	Whether the image is annotated, and when.
Annotations	Human-provided labels, shown as label : value. Includes the annotator name if available.
Predictions	AI model output, shown as label : value [confidence%]. Only predictions above the display threshold are shown. For object detection models, bounding boxes are drawn on the image.
Additional data	Extra metadata fields from the capture pipeline.
Custom tags	Tags applied to this image.
Selection button	The circle in the top-right corner. Click to select or deselect the image.

Tip

Click an image thumbnail to open the annotation view in read-only mode. This lets you examine the full-size image with its predictions and annotations.

Actions

Below the filter pane, several action buttons let you work with filtered or selected images.

Annotation actions

Button	What it does
Annotate displayed images	Creates an annotation queue from all images on the current page and opens the annotation view.
Annotate selected images	Creates an annotation queue from your selected images and opens the annotation view.
Compare selected	Opens the comparison view where you can run one or more models against the selected images.

Tag management

Below the action buttons, you can manage tags:

Control	What it does
Tag to add + Add tag	Adds a tag to all selected images.
Tag to remove + Remove tag	Removes a tag from all selected images.

What’s next?

Annotate from scratch Start labeling images when no AI model exists yet.

Understand model behavior Use filters and confidence scores to evaluate a model.

Thresholds Statistics Configure confidence thresholds for each label.

Glossary Definitions for terms used in EyVz.