This repository provides pretrained object detection models for identifying one class: “fish”.
The model was trained on the Community Fish Detection Dataset, a collaboratively built, large-scale dataset that unifies >1.9 million images and >935,000 fish bounding boxes from 17 open datasets spanning freshwater, marine, and lab environments.
With this project, our goal is to detect any fish, anywhere.
These models represent an initial training effort. They perform reasonably well across a variety of environments but can certainly be improved. If you’d like to contribute improvements or new experiments, please get in touch!
| Model | Architecture | Input image size | Target classes | Dataset | License |
|---|---|---|---|---|---|
| cfd-yolov12x-1.00.pt | YOLOv12x | 1024 | 1 (fish) | Community Fish Detection Dataset | AGPL |
git clone https://github.com/WildHackers/community-fish-detector.git
cd community-fish-detector- You can find and download the
.ptmodels from the GitHub Releases
pip install ultralyticsfrom ultralytics import YOLO
# Load model
model = YOLO("path/to/your/model.pt")
# Run on an image or folder
results = model.predict(source="path/to/images_or_videos", imgsz=1024)
# Visualize results
results[0].show()imgsz=1024); YOLO inference tools default to an image size of 640.
This model was created by a collective effort of the following folks: Filippo Varini, Dan Morris, , Sonny Burniston, Oceane Boulais, Kevin Barnard, Laura Chrobak, Alexander Merdian-Tarko, Devi Ayyagari, Mona Dhiflaoui, Joshua Chen and many others.
If you don't see your name, please email us!
We welcome further contributions!
If you:
- Have a dataset that could expand coverage,
- Want to contribute to improving the model,
...please reach out!
Below we want to provide some move visual examples that overlay the ground truth with the model detections, to give you a qualitative sense of the model's training domain.
