WingID

Model Description

A YOLO11l model fine-tuned for aircraft and bird detection. WingID is a real-time visual identification system capable of detecting and classifying flying objects — including various aircraft types and bird species — from camera feeds or static images.

Model Architecture

• Base Model: YOLO11l (Large variant)
• Framework: PyTorch / Ultralytics
• Task: Object Detection
• Input: RGB images / video frames

Training Details

• Approach: Fine-tuned YOLO11l on a curated dataset of aircraft and bird images
• Augmentations: Mosaic, random flip, scale jitter, HSV augmentation
• Optimizer: SGD / AdamW with cosine LR scheduling

Performance

Achieves high mAP on the validation set for aircraft and bird detection across multiple classes.

Files

File	Description
yolo11l.pt	Fine-tuned YOLO11l model weights

Usage

from ultralytics import YOLO
from huggingface_hub import hf_hub_download

# Download model
model_path = hf_hub_download(repo_id='devanshty/WingID', filename='yolo11l.pt')

# Load model
model = YOLO(model_path)

# Run inference
results = model('aircraft_image.jpg')
results[0].show()

Download & Use

from huggingface_hub import hf_hub_download
model_path = hf_hub_download(repo_id='devanshty/WingID', filename='yolo11l.pt')