metadata

license: apache-2.0
tags:
  - yolo
  - object-detection
  - cargo
  - packages
  - forklift
  - truck
datasets:
  - custom-cargo-package-dataset
model-index:
  - name: YOLOv8 Cargo Package Counter
    results:
      - task:
          type: object-detection
        dataset:
          name: custom-cargo-package-dataset
          type: object-detection
          split: train
        metrics:
          - type: precision
            value: 0.77187
          - type: recall
            value: 0.11111
          - type: mAP50
            value: 0.09188
          - type: mAP50-95
            value: 0.06383
          - type: F1
            value: 0.19426
language:
  - en
base_model: YOLOv8
pipeline_tag: object-detection
metrics:
  - precision
  - recall
  - f1

YOLOv8 Cargo Package Counter

This repository contains a YOLOv8-based model trained to detect and count cargo packages in images. The model was trained on a custom dataset with classe: cargo-package. It can be used for various cargo logistics and package counting tasks.

Model Description

YOLOv8 is a state-of-the-art object detection architecture, known for its speed and accuracy. This model was trained using a custom dataset containing images of cargo packages, forklifts, and trucks, making it specialized for logistics and transportation industries.

Model Architecture: YOLOv8
Number of Classes: 1 (cargo-package)
Training: The model was trained using both best.pt (the best performing model during training) and last.pt (the final checkpoint).- Use Case: Object detection and counting of cargo packages, forklifts, and trucks in warehouses, transportation hubs, or logistics centers.

Evaluation Results

The model was evaluated on the validation set using the following metrics:

Metric	Value
Precision	0.77187
Recall	0.11111
mAP50	0.09188
mAP50-95	0.06383
F1 Score	0.19426

These metrics were obtained using a threshold of 0.5 for IoU (Intersection over Union).

How to Use

You can load the model using the ultralytics library, as shown below:

from ultralytics import YOLO

# Load the model from Hugging Face
model = YOLO('https://huggingface.co./poudel/yolov8-cargo-package-counter/resolve/main/best.pt')

# Run inference on an image
results = model('path_to_image.jpg')

# Display the results
results.show()