mosquito-egg-detection

This object detection model is a finetuned version of microsoft/conditional-detr-resnet-50 model for detecting eggs (and clusters of eggs) of the Aedes aegypti mosquito.

The model was fine-tuned on the henryzord/mosquito-egg-detection dataset.

Model description

To be fulfilled.

Intended uses & limitations

To be fulfilled.

How to use

import numpy as np
from PIL import ImageDraw

import torch
from datasets import load_dataset
from transformers import AutoImageProcessor, AutoModelForObjectDetection

TOKEN = 'generate your token at https://huggingface.co/settings/tokens'

path_space = 'henryzord/mosquito-egg-detection'

image_processor = AutoImageProcessor.from_pretrained(path_space, token=TOKEN)
model = AutoModelForObjectDetection.from_pretrained(path_space, token=TOKEN)
dataset = load_dataset(path_space, token=TOKEN)
image = dataset['test'][np.random.randint(len(dataset['test']))]['image']

with torch.no_grad():
    inputs = image_processor(images=[image], return_tensors='pt')
    outputs = model(**inputs)
    target_sizes = torch.tensor([[image.size[1], image.size[0]]])
    results = image_processor.post_process_object_detection(outputs, threshold=0.3, target_sizes=target_sizes)[0]

for score, label, box in zip(results['scores'], results['labels'], results['boxes']):
    box = [round(i, 2) for i in box.tolist()]
    print(
        f"Detected {model.config.id2label[label.item()]} with confidence "
        f"{round(score.item(), 3)} at location {box}"
    )

draw = ImageDraw.Draw(image)

for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
    box = [round(i, 2) for i in box.tolist()]
    x, y, x2, y2 = tuple(box)
    draw.rectangle((x, y, x2, y2), outline="red", width=1)
    draw.text((x, y2 + 2), model.config.id2label[label.item()], fill='red')

image.show()

Training data

The model was fine-tuned on the henryzord/mosquito-egg-detection dataset. It was manually annotated by a team of students from Federal University of Santa Maria.

Training procedure

Available at this notebook.

Training hyperparameters

• num_train_epochs: 100
• learning_rate: 1e-05
• auto_find_batch_size: True

Framework versions

• Transformers: 4.42.1
• Pytorch: 2.3.1
• Datasets: 2.20.0
• Tokenizers: 0.19.1
• Timm: 1.0.7

BibTeX entry and citation info

@misc{
    author={Fulfill me},
    title={Fulfill me},
    year={2024},
    howpublished={
        Available at \url{
            henryzord/mosquito-egg-detection](https://huggingface.co/henryzord/mosquito-egg-detection
        }
    }
}