Differences
This shows you the differences between two versions of the page.
|
vss:competition:1 [2024/07/09 20:41] andy_eduard.catruna [Competition] |
vss:competition:1 [2026/07/06 18:11] (current) andy_eduard.catruna |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| ====== Competition ====== | ====== Competition ====== | ||
| - | |||
| - | The competition is hosted on Kaggle at this **[[https://www.kaggle.com/t/ca9c55b30f5e0d7e5b21d84d857a4819 | link]]**. | ||
| - | |||
| - | The teams will be composed of 2 people. | ||
| - | |||
| - | We provide you with a **[[https://www.kaggle.com/code/andreiniculae/starting-code | Starter Code]]** which demonstrates how to read the data, train a network and make a submission. You are encouraged to start your work from this notebook. | ||
| - | |||
| - | ===== Description ===== | ||
| - | |||
| - | Traditional image classification models heavily rely on accurately labeled data for training, but in real-world scenarios, acquiring large quantities of labeled images can be costly and time-consuming. | ||
| - | |||
| - | Additionally, label noise or mislabeling can further complicate the training process, leading to decreased model performance. | ||
| - | |||
| - | In this challenge, we provide you with a dataset that poses both obstacles: a significant portion of the training data remains unlabeled, and an unknown percentage of the labeled data contains mislabeled instances. | ||
| - | |||
| - | Your task is to develop innovative deep learning algorithms and techniques to overcome these challenges and build a robust image classification model. | ||
| - | |||
| - | To succeed in this competition, participants are encouraged to explore semi-supervised learning methods that leverage the unlabeled data to improve the model's performance. Developing strategies to mitigate the impact of mislabeled instances and enhance the model's ability to generalize effectively will be crucial. We encourage creative ideas. | ||
| - | |||
| - | ===== Data ===== | ||
| - | |||
| - | * 30 distinct classes. | ||
| - | * 15,000 images for training. | ||
| - | * 80% of the training set is unlabeled. | ||
| - | * an unknown percentage of samples are mislabeled. | ||
| - | |||
| - | ===== Deadline ===== | ||
| - | You can make submissions until 3:55 PM Friday (12 July). | ||
| - | |||
| - | ===== Rules ===== | ||
| - | |||
| - | <note important> | ||
| - | * Using pretrained models is not allowed. | ||
| - | * Using additional training data apart from the data provided is not allowed. | ||
| - | * Searching on the internet for the clean dataset or labels for the test set is not allowed. | ||
| - | </note> | ||
| - | |||
| - | <note important> | ||
| - | Slides | ||
| - | * Max 2 slides | ||
| - | * Use Google Slides | ||
| - | * Submit the link to your slides **[[https://forms.gle/gorF3d8GJLFkLhAs9|here]]** | ||
| - | </note> | ||
