# AITxDIL - Project Jupyter — Data

来源：https://www.kaggle.com/competitions/aitxdil-project-jupyter/data

采集时间：2026-10-19T10:05:17Z

The data given is a set of images consisting of text with background noise. The task is to use these images and denoise them using autoencoders or deep learning models of your preference. The dataset is divided into the following:
Input - Noisy Images with multiple resolutions
Output - Denoised Images with original resolutions
Submission - Unpacked individual pixel values with ID to correlate pixel locations.

**1. Training_Data.zip** - Set of 100 noisy images to be used as part of model training
**2. Cleaned_Training_Data.zip** - Set of 100 cleaned images corresponding to images in Training Dataset
**3. Test_Data.zip** - Set of 44 images to be cleaned for scoring
**4. sampleSubmission.csv** - Sample of submission file with id and value columns only