Stinger Inspector

Use Computer Vision and Machine Learning to Reclassify Cnidarians

Hackathon Findings

Hackathon Projects

• Jellyfish Bloom by Jellyfish Bloom and their presentation

• Sony Stinger Inspector repo 1 repo 2 by Sony

• Stinger Analysis by Team N and their presentation

• Finding Nematocyst by Team S

• Anemone Nematocyst Search repo 1 repo 2 repo 3 by Team Search and their presentation

• UBS Stinger Classification by Team W and their presentation

Background

The world's oceans cover two-thirds of the planet's surface, and as you'd expect for such a vast area, they contain multitudes of an unbelievable variety of life. The phylum Cnidaria is extremely diverse in form (morphology), which belies their simplistic structural organization; they are primarily a matrix of cellular tissues, and do not have any kinds of organs or systems at all! Jellyfish, corals, anemones are all examples of Cnidarians.

Every type of Cnidarian has nematocysts

Capsules Cnidarians (jellyfish, corals, sea anemones) Unifying characteristic cellular products that are very complicated "stinging cells" - they are NOT cells they are cellular products 3 types: sticky, penetrating (current focus, nematocysts - most diverse, w/wout venom), enveloping

A diagram showing the structure of the stinging nematocyst capsule

"eversible" - explosive, discharge-able - fastest cellular mechanism know (700 ns) response to pressure, not necessarily responding to threat - for prey and for defense

How they open is different

5 - 100 microns in size

maturity of capsule - does the shaft change length/shape within over time location of capsule - filament, pharynx shape of capsule - ovoid, thicker vs. slender, transparency of capsule - more or less opaque features of shaft of tubule length of shaft shapes of shaft ("V" shape or forked)

species code

LAS Leica software for indicating on microscopes

---

Solutions

Given how little we know about Cnidarians, and how important coral reefs and marine ecosystems are, there are surprisingly few online resources for Cnidaria.

Machine learning to help with classifying cnidarians. Reclassify images from literature. Possibly use crowdsourcing to help add "objectivity" Library API - Get list of journals for specific nematocyst / cnidarian related articles. Other online data sources to get additional images, Identify other publications that have nematocyst articles that can be collected: 1) Must identify the capsule within the image; 2) Must orient the capsule to the same direction.

Some possible solutions might be:

• A solution involving a thing. What's the best way to make this thing?

• Anything else you can think of!

---

Resources

• AMNH Cnidarian collections: The AMNH cnidarian database
• AMNH SciCafe "Amazing Anemones" video: SciCafe with Estefania Rodriguez
• YouTube: Nematocysts Firing
• Hexacorallians of the World:
• Cnidaria page at "Tree of Life":
• Biogeoinformatics of Hexacorals: Geospatial, taxonomic, and environmental data for a group of marine invertebrates.

---

Challenge owner: Estefania Rodriguez and Luciana Gusmão