Metal Band Logos Classification

Summary

• Introduction
• Parameters
• Examples
• Usage for developers
• Reference
• License

Introduction

Metal Band Logos Classification is a task that involves the classification of metal band logos into different categories or classes. The goal is to train a model that can accurately identify and categorize logos associated with various metal bands.

For this task, the fastai library is utilized. fastai is a high-level deep learning library built on top of PyTorch, providing a simplified and intuitive interface for training and deploying deep learning models.

Using the fastai library for Metal Band Logos Classification simplifies the process of building and training a classification model. The library provides pre-built functions and classes that handle common tasks such as data preprocessing, model architecture selection, and training loop implementation.

In this task, there are common labels such as black-metal and epic-metal.

Parameters

Inputs

• input - (image -.png|.jpg|.jpeg): the input is an image containing a metal band logo. This image can be collected from various sources such as websites, music albums, t-shirts, or any other sources that contain metal band logos

Output

• output - (text): Includes the model's predictions for each label, displayed as probabilities.

Examples

input	output
	{black-metal: 0.9999315738677979, epic-metal: 6.84056431055069e-05}

Usage for developers

Please find below the details to track the information and access the code for processing the model on our platform.

Requirements

python~=3.10
torch~=2.0.0
fastai

Code based on AIOZ structure

from fastai.vision.all import *
import torch, os

...
def do_ai_task(
        input: Union[str, Path],
        model_storage_directory: Union[str, Path],
        device: Literal["cpu", "cuda", "gpu"] = "cpu",
        *args, **kwargs) -> Any:
        
        model_id = os.path.abspath(model_storage_directory + "...")
        learn = load_learner(model_id)
        if torch.cuda.is_available():
            learn.model.cuda()

        labels = learn.dls.vocab

        img = PILImage.create(input)
        pred,pred_idx,probs = learn.predict(img)
        output = {labels[i]: float(probs[i]) for i in range(len(labels))}

    return output

Reference

This repository is based on and inspired by Oriol's work. We sincerely appreciate their generosity in sharing the code.

License

We respect and comply with the terms of the author's license cited in the Reference section.