You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

VehicleNet-Y26 License & Usage Terms

VehicleNet-Y26m is a vehicle detection model released under the Apache License, Version 2.0. Access to this model is permitted only to individuals of legal age who have the authority to accept and comply with the license terms in their jurisdiction. By accessing, downloading, or using VehicleNet-Y26, you confirm that:

You meet the legal age requirements applicable in your country.
You have the authority to accept this agreement.
You agree to use the model in compliance with the Apache-2.0 license and all applicable laws and regulations.
VehicleNet is provided as-is, without warranties of any kind, express or implied. The authors make no guarantees regarding accuracy, reliability, fitness for a particular purpose, or suitability for deployment in safety-critical or regulated environments.
You are solely responsible for how the model and its outputs are used. Any misuse, unlawful application, or deployment without proper validation is strictly discouraged.
If you do not agree to these terms or lack the authority to accept them, do not use this model.

VehicleNet-Y26m

VehicleNet-Y26m is another multi-class vehicle detection model designed for fine-grained vehicle type recognition in real-world traffic scenes. The model is trained on UVH-26-MV Dataset released by IISc Banaglore. The dataset is based on Indian traffic which is highly challenging, dense and heterogeneous. It contains 14 vehicle categories such as hatchback, sedan, SUV, MUV, two-wheelers, three-wheelers, buses, trucks, and commercial vehicles. This m variant is designed for speed and inferences on low-latency devices, offering significant speed and accuracy. This model is finetuned on YOLO26m:arXiv model by Ultralytics using UVH-26-MV Dataset.

Model Overview and Parameters

Pretrained_weights: YOLO26m
Number of Classes: 14
Layers: 132 layers
Parameters(M): 20,360,246 parameters, 0 gradients
GFLOPs: 67.9
Input Resolution: 640 × 640
Training Epochs: Up to 60 (early stopping applied, patience=5), best model at: 35/60
Batch Size: 48
Hardware: Dual NVIDIA Tesla T4 GPUs
Framework: Ultralytics YOLO (PyTorch)

Performance Summary

mAP@50: 0.74967
mAP@50:95: 0.6685
Precision: 0.70126
Recall: 0.71083

Per-class mAP@50:95

The model showed strong detection performance for structurally distinct vehicle categories such as two-wheelers, three-wheelers, buses, and trucks. Fine-grained car subclasses (hatchback, sedan, SUV, MUV) exhibit expected inter-class confusion/challenge due to visual similarity and viewpoint overlap, as reflected in the confusion matrix.

Intended Use

The model is suitable for:

Edge device computation
Traffic surveillance and analytics
Academic research and benchmarking

License

This model is released under the Apache License 2.0.

Downloads last month: 1

Model tree for Perception365/VehicleNet-Y26m

Base model

onnx-community/yolo26m-ONNX

Finetuned

(1)

this model

Perception365
/

VehicleNet-Y26m