> For the complete documentation index, see [llms.txt](https://docs.unitlab.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.unitlab.ai/documentation/multimodal-annotations/media-and-document-layouts.md).

# Media and document layouts

Put videos, images, sensor logs, PDFs and text of one case in a Data Group and label each file with its own tools.

Some decisions need several kinds of evidence at once: a video and the work instruction it follows, a photo and its inspection report, a recording and its transcript. A Data Group shows these files as tiles in one Workbench. Each tile keeps its own viewer and tools.

![A smartwatch test episode: head-camera video, still frame, test pad sensor chart and work instruction PDF](https://292810646-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGjVLUz4wthGkGlRKM6rM%2Fuploads%2FITK6LgE97GUHVuTIMvev%2Fmultimodal-data-group-hero.webp?alt=media)

## Example combinations

| Case               | Files in the group                                              |
| ------------------ | --------------------------------------------------------------- |
| Robot task episode | Camera video, gripper or force sensor CSV, task instruction PDF |
| Part inspection    | Part photo, thermal image, inspection report PDF                |
| Field service      | Site photo, equipment log CSV, technician's notes as text       |
| Support call       | Call recording, transcript, case document                       |

Upload each file in its own format. A PDF stays a PDF with pages, a CSV stays a chart, and a video keeps its frames.

## What each tile can do

| Tile type          | What you label there                                       | Learn more                                                                                                    |
| ------------------ | ---------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------- |
| Video              | Tracked boxes, polygons, masks and actions on the timeline | [Video annotation](https://docs.unitlab.ai/documentation/annotations/video-annotation.md)                     |
| Image              | Boxes, polygons, masks, keypoints                          | [Image annotation](https://docs.unitlab.ai/documentation/annotations/image-annotation.md)                     |
| CSV as time series | Intervals and instants on one channel or the whole file    | [Sensor annotation](https://docs.unitlab.ai/documentation/annotations/sensor-annotation.md)                   |
| PDF or document    | Boxes and other shapes on each page                        | [Document & PDF annotation](https://docs.unitlab.ai/documentation/annotations/document-and-pdf-annotation.md) |
| Text               | Entities and relations                                     | [Text annotation](https://docs.unitlab.ai/documentation/annotations/text-annotation.md)                       |
| Audio              | Segments on the waveform                                   | [Audio annotation](https://docs.unitlab.ai/documentation/annotations/audio-annotation.md)                     |

## Build a layout

{% stepper %}
{% step %}

### Name the files by case

Give every file of a case the same ID, for example `ep01_head_camera.mp4`, `ep01_station_log.csv` and `ep01_work_instruction.pdf`.
{% endstep %}

{% step %}

### Add Auto-Groups

In **Data Assets**, click **New Folder** > **Add Auto-Groups**. In **Configure** and **Rules**, choose the folder, add one tile per role (camera, sensor, instruction) and set which tiles are required.
{% endstep %}

{% step %}

### Arrange the tiles

In the **Layout** step, pick **Grid** (all files in a balanced grid), **List** (one main tile, the rest below) or **Custom**. In **Custom**, drag tiles into the workspace to place and resize them, and click **Adjust** to fill empty space with even gaps. Panels can't overlap. Give the main evidence the largest tile.
{% endstep %}

{% step %}

### Attach and set up CSV files

Attach the grouped folder to a project. If a group has a CSV, its card shows **Needs configuration**. Click it, choose **Time series** or **Tabular**, pick the X axis and the channels, then click **Import time series**.
{% endstep %}
{% endstepper %}

![The CSV setup dialog with Time series selected, time\_s as X axis and two channels](https://292810646-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGjVLUz4wthGkGlRKM6rM%2Fuploads%2FyIRoGE4QCbsCpWaw8woh%2Fmultimodal-csv-configuration.webp?alt=media)

## Label a case

1. Open the group. The first tile is active.
2. Click the tile you want to label. The toolbar, the class list and the timeline change to that file's type.
3. Label it, then click the next tile.
4. Watch the tile headers: a green check means saved and valid, **Invalid** means a required value is missing.
5. Click **Send to Review** when the required tiles are done.

![With the sensor tile active, the toolbar shows the Charging class and the timeline lists the Charging intervals](https://292810646-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FGjVLUz4wthGkGlRKM6rM%2Fuploads%2F4fPh2NRKer4gsh54zCAp%2Fmultimodal-active-panel-sensor.webp?alt=media)

## Tips

* Classes are shared by the project. Add only the geometry each file needs: an **interval** class for the CSV, a **box** class for the PDF.
* A label on one tile is not linked to a label on another. If reviewers need the link, record it in a property, such as the step number.
* Check one finished case with a reviewer before you label the rest. Fix the tile order and sizes then, not after a hundred cases.

## Related

* [Multimodal overview](https://docs.unitlab.ai/documentation/multimodal-annotations/multimodal-overview.md) · [Data Groups and layouts](https://docs.unitlab.ai/documentation/data/data-groups-and-layouts.md) · [Multi-camera video](https://docs.unitlab.ai/documentation/multimodal-annotations/multi-camera-video.md) · [Medical study and clinical document](https://docs.unitlab.ai/documentation/multimodal-annotations/medical-study-and-clinical-document.md)
