|
--- |
|
license: agpl-3.0 |
|
base_model: |
|
- Ultralytics/YOLO11 |
|
tags: |
|
- image-classification |
|
- pytorch |
|
- yolo |
|
--- |
|
|
|
# Not Sea Otter |
|
|
|
Yes, sea otters are incredibly CUTE, but you might not actually know what a sea otter looks like. |
|
|
|
Can you distinguish between a sea otter, seal, sea dog, sea lion, walrus, manatee, and dugong? |
|
|
|
Can you recognize the differences between sea otters, groundhogs, beavers, nutrias, capybaras, mongooses, and martens' face? |
|
|
|
Can you tell apart a sea otter from a river otter? They are not the same! |
|
|
|
This model was fine-tuned on YOLO11 in order to classify these animals often confused with sea otters. |
|
|
|
The model was trained using thousands of animal images, with 80% for training and 20% for validation. It achieves an precision of 60% - 90% on the validation set, varying by animal. |
|
|
|
## usage |
|
|
|
```python |
|
from huggingface_hub import hf_hub_download |
|
from ultralytics import YOLO |
|
from PIL import Image |
|
|
|
model_path = hf_hub_download(repo_id="YvesX/not-sea-otter", filename="model.pt") |
|
|
|
model = YOLO(model_path) |
|
|
|
image_path = "/path/to/image.jpg" |
|
image = Image.open(image_path).convert("RGB") |
|
|
|
results = model(image) |
|
|
|
if results[0].probs: |
|
top_class_index = results[0].probs.top1 |
|
top_confidence = results[0].probs.top1conf.item() |
|
print(f"This is a {model.names[top_class_index]}! (with confidence: {top_confidence:.2f})") |
|
|
|
``` |
|
|
|
## label |
|
|
|
| id | label | 中文 | English | actually | sample photos | |
|
|----|----------|--------|-------------|----------------|---------------| |
|
| 0 | capybara | 水豚 | Capybara | Hydrochoerus | 1851 | |
|
| 1 | diao | 貂 | Marten | Martes | 2669 | |
|
| 2 | haibao | 海豹 | Seal | Phocidae | 4397 | |
|
| 3 | haigou | 海狗 | Sea Dog | Callorhinus | 134 | |
|
| 4 | hainiu | 海牛 | Manatee | Trichechidae | 1118 | |
|
| 5 | haishi | 海狮 | Sea Lion | Eumetopias | 893 | |
|
| 6 | haita | 海獭 | Sea Otter | Enhydra | 1552 | |
|
| 7 | haixiang | 海象 | Walrus | Odobenidae | 198 | |
|
| 8 | heli | 河狸 | Beaver | Castor | 4673 | |
|
| 9 | helishu | 河狸鼠 | Nutria | Myocastor | 3963 | |
|
| 10 | meng | 獴 | Mongoose | Herpestidae | 2083 | |
|
| \* | | | | | | |
|
| 12 | rugen | 儒艮 | Dugong | Dugongidae | 223 | |
|
| 13 | shuita | 水獭 | River Otter | Lutra | 904 | |
|
| 14 | tuboshu | 土拨鼠 | Groundhog | Marmota | 3097 | |
|
|
|
|
|
\* In Chinese, the pronunciation of "海豹" (seal) is identical to "海报" (poster). For humor somehow, the model also includes a `poster` label. ~~Movie poster recognition is quite accurate.~~ |
|
|
