Context

Semantic Segmentation is one of major tasks in Computer Vision. It is the pixel-wise classification of an image into object classes. This dataset contains 1000 images and segmentation masks pairs of individual people's clothing. With 59 object classes and a relatively lesser data, the task of modelling is expected to be a challenging one! The data needs no preprocessing, all images are of same size, same format, and ready to model.

Content

The dataset contains 1000 images and 1000 corresponding semantic segmentation masks each of size 825 pixels by 550 pixels in PNG format. The segmentation masks belong to 59 classes, the first being the background of individuals, and the rest belong to 58 clothing classes such as shirt, hair, pants, skin, shoes, glasses and so on. A CSV file containing the list of 59 classes is included in the dataset. The dataset contains data in both JPEG formats and PNG formats. However, JPEG is found to be lossy, while PNG is lossless with the essence of Originality.

Acknowledgements

The raw contents of this dataset are collected as such from the Github Repository: https://github.com/bearpaw/clothing-co-parsing. However, this dataset is a transformed version of the original dataset. The original work can be found at: @inproceedings{yang2014clothing, title={Clothing Co-Parsing by Joint Image Segmentation and Labeling}, author={Yang, Wei and Luo, Ping and Lin, Liang} booktitle={Computer Vision and Pattern Recognition (CVPR), 2014 IEEE Conference on}, year={2013}, organization={IEEE} }

Inspiration

The original dataset acknowledged above has images of different sizes and different formats such as JPG and MATLAB. This kernel https://www.kaggle.com/rajkumarl/how-to-convert-matlab-file-into-png-image-file processed the above mentioned original dataset and transformed every data into a ready-to-use format of uniform size. No image or semantic segmentation mask available in this dataset is a copy of original dataset. This dataset is a completely transformed version, prepared to help Computer Vision people concentrate on their modeling part, rather than spend time preprocessing the images and masks.