---
title: OceanCV FirstPass
emoji: 🦑
colorFrom: blue
colorTo: indigo
sdk: streamlit
app_file: app.py
pinned: false
language: en
license: mit
tags:
- ocean
- computer-vision
- yolo
- detection
- marine-science
datasets:
- OceanCV
metrics:
- map
- precision
- recall
---

# OceanCV_FirstPass

Unified marine object detection model.

### Dataset Overview
- **Total Images**: 8,605
- **Total Annotations**: 85,557
- **Classes**: 1 (object)
- **Input Resolution**: 1024x1024

### Performance Metrics (Epoch 87)
| Metric | Value |
| :--- | :--- |
| **mAP50** | 85.1% |
| **mAP50-95** | 61.3% |
| **Precision** | 87.7% |
| **Recall** | 77.8% |

### Visualizations
![Results](results.png)
![Confusion Matrix](confusion_matrix.png)
![Normalized Confusion Matrix](confusion_matrix_normalized.png)
![F1 Curve](BoxF1_curve.png)
![Precision Curve](BoxP_curve.png)
![Recall Curve](BoxR_curve.png)
![PR Curve](BoxPR_curve.png)
![Labels](labels.jpg)

### Usage
```python
from ultralytics import YOLO
model = YOLO("OceanCV_FirstPass.pt")
results = model.predict("image.jpg", imgsz=1024)
```