Cameron Trotter

COME WORK WITH ME!!!

Very exciting open faculty position at MIT, with ecosystem and biodiversity modeling as a focus!

cee.mit.edu/people/share...

The academic starter pack.

Thanks to the Technology Specialist Network’s International Conference Fund for supporting my travel to ICCV!

A photo of me (left of the image) standing at a lectern giving my talk. My title slide is to the right of me, entitled “Automated detection of Antarctic benthic organisms in high resolution in situ imagery to aid biodiversity monitoring”.

Me (right) next to my massive poster (left). The poster introduces my research.

An image of another talk during the main conference showing the use of my Northumberland Dolphin Dataset as a benchmark for segmentation foundation models.

A view of the poster hall with ice cream freezers in front!

#ICCV25 was a blast! Presented my work on automating Southern Ocean benthic monitoring at the Marine Vision workshop & joined the AI for Conservation bird walk. I loved seeing my Northumberland Dolphin Dataset still used for benchmarking segmentation models! Lots of collab ahead! 📄 bit.ly/4qItEdQ

The CVPR 2026 Workshop Proposal site is now live on OpenReview! 🎉

Submit your ideas and help shape the future of computer vision and AI:
openreview.net/group?id=the...

@cvprconference.bsky.social

Badass birders braving the rain this morning!! We got soaked, saw cool birds, and got a beautiful rainbow as a reward 🌈

Artificial Intelligence supercharges science on the Antarctic seafloor - British Antarctic Survey Using the latest AI technology, scientists have dramatically sped up the process of detecting animals found in photographs and video taken of the seafloor.

🌟 ‘Automated Detection of Antarctic Benthic Organisms to Aid Biodiversity Monitoring’ is by Cameron Trotter, Huw Griffiths, Tasnuva Ming Khan + Rowan Whittle.

It was published at the Joint Workshop on Marine Vision, at the International Conference on Computer Vision ⬇️

www.bas.ac.uk/media-post/a...

Image 2 is a slightly more busy image of the sea floor. You can easily see there's a variety of shelled creatures and corals.

Image two is labeled and shows that there's over 100 creatures in the shot.

📣 Dr @camerontrotter.bsky.social explains:

"The images are packed full of weird and unusually shaped animals, often living on top of each other - some of which have never been seen before."

The team trained AI to identify seafloor creatures near Antarctica, like starfish, corals, and fish.

🧵 2/4

Image 1 is a pretty empty image of the Antarctic Sea floor. There's visibly a few large starfish in bright pink.

In the annotated version of image 1, there's around 15 creatures labelled, including some tiny shells scattered across the floor.

Supercharged seafloor spotting! 🐚

Our AI and science teams have been using machine learning to label surveys of the Antarctic seafloor. It's dramatically sped them up from <8 hours per photo, to a few seconds.

This one looks like a quick job. But stay with us as we up the difficulty...

🧵 1/4

Sea Otters Are Stealing Surfboards in California. Again.

www.nytimes.com/2025/10/18/s...

An unknown sea otter was spotted stealing surfboards at Steamer Lane in Santa Cruz this week. Is 841 back to her old ways or have other otters learned to hang 10? Check out my latest for @nytimes.com

One way sighted people can determine what to put in alt text: Imagine you sent the post or article to a friend, and the image didn't load. What would someone need to know to get the equivalent experience and context? This will help you identify the important details.

Are you coming to @iccv.bsky.social??

Do you like BIRDS?!?! The #CVBirdWalk is happening!!!

Oct. 21, 7am, Ala Moana Park. Kick off your conference right!!

#ICCV2025 #AIforConservation

Interested in doing a PhD in machine learning at the University of Edinburgh starting Sept 2026?

My group works on topics in vision, machine learning, and AI for climate.

For more information and details on how to get in touch, please check out my website:
homepages.inf.ed.ac.uk/omacaod

Figure 2 from the linked paper, showing our methodology workflow. We take the seafloor training imagery and split it based on substrate, depth, and seafloor angle. This is then patched down and processed using SAHI, which allows us to train an object detection model on a patch-level, retaining resolution. We then postprocess the detections, reprojecting them back to their location on the original image.

Heading to Honolulu for #ICCV? 🌺 Ease into the week with my talk on automating detection of seafloor organisms around Antarctica for biodiversity monitoring! 🐙 Catch it at the Joint Workshop on Marine Vision, Oct 19, Room 318B, 09:50–10:20. 📄: arxiv.org/abs/2507.21665

Welcome to #Plymouth, @bas.ac.uk 🌊

The #RRSDavidAttenborough arrived here safe and sound yesterday, and we're excited to share our team will be heading on board later today. Stay tuned for updates!

📹 Filmed by PML's Jani Pewter, from our lab roof!

#BoatyMcBoatface #PolarScience #MarineScience

A screenshot of the Google scholar metrics for my profile, showing I have achieved 100 citations.

🎉 ECR milestone achieved: My research has now been cited 100 times! 🎉

A photo showing myself (top-left) standing in front of a room of people (4 in shot) giving a talk. I am standing next to a presentation slide (top-right) displaying 4 QR codes directing to our research paper, code, dataset, and guide.

Great time at this year’s NERC Digital Gathering, chatting with colleagues from across NERC-funded centres and sharing ideas about the role of emerging technologies in environmental science.

Lots of interesting discussion from my talk on Southern Ocean benthic image analysis using computer vision!

Congratulations to everyone who got their @neuripsconf.bsky.social papers accepted 🎉🎉🎉

At #EurIPS we are looking forward to welcoming presentations of all accepted NeurIPS papers, including a new “Salon des Refusés” track for papers which were rejected due to space constraints!

This is the same for the UK. I wouldn’t be allowed to book an Airbnb _or_ a conference hotel. Ignoring some nuance, I simply have to take the cheapest option available to me through our contracted travel agent. I do not book hotels or travel myself and claim back, it’s done for me.

British Antarctic Survey Careers | Seabird Remote Sensing Data Analyst Interview date – 16 September 2025

We are hiring:
bas.ciphr-irecruit.com/applicants/v...

Big thanks to my coauthors @huwiceandstuff.bsky.social, @ming-tfk27.bsky.social, and @rowanwhittlebas.bsky.social for their support!

@bas.ac.uk

🧵4/4

🌺 I'll be presenting the work in person this October in Honolulu, Hawaii.

Come say hi at #ICCV2025 if you're around, especially if you're interested in collaborations in the marine vision space!

🧵3/4

An overview diagram showing how our new framework works. We start by taking in large scale benthic images as input, splitting them into training, validation, and testing sets based on substrate, depth, and sea floor angle. Then we use SAHI to patch the high res images down to small image chunks. An object detection model is then trained using the patches. We perform some post processing, generating a large high-res image with detections as output.

We

🪸Tackle key challenges in marine imagery: scarce annotations, variable object sizes, & complex seafloor terrain

📷 Combine resolution-preserving patching, spatial augmentation, & SAHI to detect 25 key morphotypes

❄️ Release the first CV dataset of Antarctic benthic organisms

🧵2/4

An example output from our framework. An image of the Antarctic seafloor, with hundreds of organisms on it. The organisms each have a bounding box around them, a class label, and a confidence score produced by our framework.

🇦🇶 Our new paper detailing the automated detection of Antarctic benthic organisms using high-res in situ imagery has been accepted to the #ICCV2025 Joint Workshop on Marine Vision!

📄 Read the preprint now: arxiv.org/abs/2507.21665

🧵1/4

EurIPS is coming! 📣 Mark your calendar for Dec. 2-7, 2025 in Copenhagen 📅

EurIPS is a community-organized conference where you can present accepted NeurIPS 2025 papers, endorsed by @neuripsconf.bsky.social and @nordicair.bsky.social and is co-developed by @ellis.eu

eurips.cc

Join us at the Joint Workshop for Marine Vision at #ICCV25

We have full-paper and nectar tracks.
Accepted papers are included in the official ICCV workshop proceedings.
Submission deadline: June 30

More at: vap.aau.dk/marinevision/

@iccv.bsky.social
@jbhaurum.bsky.social
@justin-kay.bsky.social

A view of Amara station in San Sebastián at night.

A map of Western Europe showing my journey. The red marker is Newcastle, and the blue circle is San Sebastián.

And I’ve made it! About 14h from Newcastle to San Sebastián via London, Paris, and Hendaye, but that wasn’t much longer than flying (no direct) and *far* more environmentally friendly! Now to get some sleep ready for tomorrow!

Thanks to @seatsixtyone.bsky.social for providing my route!

A view of the entrance to the light rail station in Hendaye. It’s a small black stone building with the words “Euskotren Hendaia” above the door.

A view of the Euskotren to San Sebastián, at the station in Hendaia.

5h later and I’ve arrived in Hendaye/Hendaia on the 🇫🇷/🇪🇸 border! Now just a quick light rail to San Sebastián.

Inkscape plugin to rescale figures without distorting the text, and other useful features!
github.com/burghoff/Sci...

A view from the Eurostar platform at Gare du Nord after arriving in Paris.

An image of Gare du Nord station, Hall 1. On the Mezzanine level above is a sign stating “Hall 2” with a Union Jack, and an arrow pointing left. Behind a group of people in the middle of the scene is a sign pointing down for the metro.

Gare Montparnasse from the outside; a mixture of concrete and glass.

Boarding the TGV train to Hendaye from Gare Montparnasse. The train is to the right of the image. Top left, an electronic board shows the route: Bordeaux Saint-Jean, Dax, Bayonne, Biarritz, Saint-Joan-de-Luz-Ciboure, and Hendaye.

Next, a very quick metro from Gare du Nord to Gare Montparnasse, ready to catch a TGV across France to Hendaye, on the French-Spanish border.