Skip to content

🏷️ Tags

The tags in Panoramax are an extra layer of semantics useful to:

  1. describe all pictures linked to a sequence
  2. give more details associated to a picture
  3. describe in detail a specific object within a picture

By principle, there is no hard restriction on how tags may be created, edited and deleted.

Where you can find tags

Sequences

It is possible to give context for all pictures of a specific sequence. For example, the upload page allows an easy selection of the transport mode:

Sequence transport mode selector

In the upload page, a dedicated field allows to add any tag you want.

Pictures attributes

The pictures tags allows to give a bit more context about the picture capture condition, or what kind of object it can show.

Tip

This is not to be confused with pictures metadata that is read from EXIF or XMP tags. To know more about pictures metadata, you can read our dedicated documentation.

This is useful for example if you have a classification model that can say "there is a recycling container in this picture", but not exactly where in the picture.

Picture tag

Objects in a picture

You can draw precisely in a picture the position of an object you'd like to describe. This is very similar to annotations you can set for training detection models.

Objects in a picture

📖 Tags syntax

Info

Read more about detailed tags syntax in the reference page.

A few examples of tags you can use:

transport=car
osm|highway=street_lamp
wd|P31=Q125191
detection_model[osm|traffic_sign=yes]=SGBlur-yolo11s/0.1.0
detection_confidence[osm|traffic_sign=yes]=0.875

✏ Use and edit tags

With web viewer

The simplest way to access existing tags is through the web viewer, in a picture details widget, in Tags tab:

Tags details tab

If you are logged-in on your Panoramax instance, you can edit them as well:

Edit picture tags

Via API

Our API offer all necessary endpoints to work with tags and semantics. You can find the API reference here. Some useful calls:

  • GET/PATCH /api/collections/:cid/items/:iid: to get or edit semantics of a picture
  • POST /api/collections/:cid/items/:iid/annotations: create a new annotation on a picture
  • GET /api/search: the filter search parameter allows to look for pictures having a certain tag defined (example)

With Prolix CLI

The Prolix tool aims to simplify detecting, classifying and share features visible on Panoramax pictures. As now, it works for traffic signs in France, but feel free to come and help us expand its capabilities!