scizzum/model_10_22_run

EVA-02 Giant Roadwork Detector

Fine-tuned EVA-02 Giant (eva_giant_patch14_224.clip_ft_in1k) model for roadwork detection

Model Details

• Architecture: EVA-02-Giant (eva_giant_patch14_224.clip_ft_in1k)
• Task: Binary image classification (Roadwork detection)
• Training Accuracy: 99.20%
• Framework: timm (PyTorch)
• Input Size: 224x224
• Number of Parameters: ~1B

Usage

import timm
import torch
from PIL import Image
from torchvision import transforms

# Load model
model = timm.create_model('eva_giant_patch14_224.clip_ft_in1k', pretrained=False, num_classes=2)
model.load_state_dict(torch.load('pytorch_model.bin'))
model.eval()

# Prepare image
transform = transforms.Compose([
    transforms.Resize(224),
    transforms.CenterCrop(224),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.48145466, 0.4578275, 0.40821073],
                       std=[0.26862954, 0.26130258, 0.27577711])
])

image = Image.open('your_image.jpg')
input_tensor = transform(image).unsqueeze(0)

# Inference
with torch.no_grad():
    output = model(input_tensor)
    prediction = torch.nn.functional.softmax(output, dim=1)
    
print(f"No Roadwork: {prediction[0][0]:.2%}")
print(f"Roadwork: {prediction[0][1]:.2%}")

Classes

• 0: No Roadwork
• 1: Roadwork

Submitted By

5Cq7fjH5kobu65GJ8gvK9hh7TY6d3M4hi7gjvcv4sk

Submission Time

2025-10-23 14:11:51