Skip to content

Welcome to finwave

  • What finwave is and what problem it solves
  • Who it’s for
  • The key concepts you’ll see again and again
  • Where to go next

finwave is a platform for wildlife photo-identification — the practice of recognising individual animals from photos of their natural markings (dorsal fin nicks, fluke patterns, body scars, eye-patches, and so on).

It is built around three things:

  • Catalogues of individuals — every recognised animal in a population gets a profile with images, sighting history, and relationships to other individuals.
  • Encounters — every time an observer photographs a group of animals in the field, the resulting set of photos is recorded as an encounter. Encounters are the raw observational record everything else flows from.
  • Identifications — the link between an individual in the catalogue and a photo in an encounter. finwave uses ML models to propose identifications, but every identification is tracked with an evidence tier so you always know how strong the evidence is.
  • Researchers running population studies (mark-recapture, social network, predator-prey, distribution).
  • Conservation organisations maintaining long-running catalogues for a population.
  • Citizen-science contributors submitting opportunistic encounters to a population they’ve been invited to.
  • Field photographers who want their photos turned into useful long-term observational records.

If you’re not sure which of these applies to you, start anyway — finwave’s permission model adapts to your role inside each population.

ConceptOne-line summary
PopulationA bounded study group (e.g. “Southern Resident Killer Whales”). Almost everything in finwave is scoped to a population.
EncounterOne field observation: photographer + location + date + photos.
IndividualA recognised animal with a profile, sighting history, and (optionally) social relationships.
AnnotationA bounding box on a photo identifying which animal is shown.
Evidence tierHow strong the evidence behind an identification is — Attested, ModelSuggested, ModelCorroborated, or Confirmed.
OrganisationThe team or institute under which an encounter was submitted. Controls visibility of private data.
RoleYour permission level inside a population — Novice, Expert, Professional, or Administrator.

You don’t need to memorise these now. They are linked from every page that uses them.

  1. Quick Start — create an account, join a population, and submit your first encounter.
  2. Navigating the App — what each section of the sidebar does.
  3. Core Concepts — read in order if you’d like a guided tour.

If you have a backlog of thousands of historical photos sitting in a folder or cloud bucket, also see the Desktop Client — it is the right tool for bulk onboarding.