Instructions to use TommyPanLab/traffic-violation-yolov8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use TommyPanLab/traffic-violation-yolov8 with ultralytics:
from huggingface_hub import hf_hub_download from ultralytics import YOLO # pick the weights file from this repo's "Files and versions" tab weights = hf_hub_download("TommyPanLab/traffic-violation-yolov8", "<weights>.pt") model = YOLO(weights) source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
Traffic Violation YOLOv8 Weights
這個 repository 儲存交通違規偵測專案使用的自訂 YOLOv8 權重與 DeepSORT ReID checkpoint,搭配 GitHub 原始碼專案使用:
Tommyweige/Yolo-Pedestrian-Detection-for-Right-of-way-Recognition
目前 repository 保持 private。模型權重的資料集來源、衍生權重發布權與 第三方元件授權仍需由擁有者確認;在確認前不應改為 public 或重新散布。
權重內容
| 路徑 | 用途 | 類別 |
|---|---|---|
traffic-light/yolov8s_tf.pt |
traffic-light | car, motorbike, traffic light, zebra, truck, bus |
traffic-light/yolov8l_tf.pt |
traffic-light | bus, car, motorbike, traffic light, truck, zebra |
traffic-light/yolov8x6_tf.pt |
traffic-light | bus, car, motorbike, traffic light, truck, zebra |
right-of-way/yolov8s_zebra.pt |
right-of-way | bus, car, motorbike, person, truck, zebra |
right-of-way/yolov8l_zebra.pt |
right-of-way | bus, car, motorbike, person, truck, zebra |
right-of-way/yolov8x6_zebra.pt |
right-of-way | bus, car, motorbike, person, truck, zebra |
tracking/ckpt.t7 |
DeepSORT ReID checkpoint | — |
模型檔案的大小與 SHA-256 會記錄在 GitHub 專案的
model_manifest.json。應用程式下載後會驗證雜湊,不符合時會停止推論。
使用方式
GitHub 專案會固定下載 revision v1.0.0,並將權重快取到使用者目錄。
也可以使用 TRAFFIC_MODEL_DIR 指向本機權重資料夾,做離線開發。
目前沒有在此卡片填寫 precision、recall 或 mAP;這些數值尚未完成可重現 的評估流程。
相關授權
請參閱 GitHub 專案的 THIRD_PARTY_NOTICES.md。這個 model repository
在權利確認前不授予公開散布或商業使用權。
- Downloads last month
- 142