AgroMind Plant Disease Classifier (MobileNetV2)

Model Description

MobileNetV2 image classifier trained on the New Plant Diseases Dataset to detect 38 plant disease classes. Serves as a lightweight fallback for the NFNet-F1 model.

Framework

• Architecture: MobileNetV2 (torchvision)
• Format: PyTorch checkpoint (.pth)
• Input size: 224×224 RGB (resize to 256, center crop to 224)
• Normalization: ImageNet (mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])

Usage

from huggingface_hub import hf_hub_download
import torch
from torchvision import models, transforms
from PIL import Image

repo_id = "Arko007/agromind-plant-disease-mobilenet"
ckpt = hf_hub_download(repo_id, "newplant_model_final.pth")
labels_path = hf_hub_download(repo_id, "labels.txt")

with open(labels_path) as f:
    labels = [l.strip() for l in f if l.strip()]

model = models.mobilenet_v2(pretrained=False)
model.classifier[1] = torch.nn.Linear(model.classifier[1].in_features, len(labels))
state = torch.load(ckpt, map_location="cpu")
model.load_state_dict(state)
model.eval()

transform = transforms.Compose([
    transforms.Resize(256),
    transforms.CenterCrop(224),
    transforms.ToTensor(),
    transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),
])

img = Image.open("leaf.jpg").convert("RGB")
with torch.no_grad():
    logits = model(transform(img).unsqueeze(0))
    print(labels[logits.argmax(dim=1).item()])

Output

Returns logits for 38 plant disease classes. See labels.txt for class names.