aRTF Clothes dataset

<p>A dataset of almost 2000 RGB images of clothes in real-world household scenes. All images* are annotated with semantic keypoints, object masks and object bounding boxes using the <a href="https://cocodataset.org/#format-data">COCO keypoints </a>format.&nbsp;</p> <p>&nbsp;</p> <p>The table below summarizes the dataset. Note that the scenes and cloth items for the train and test split are completely distinct to measure generalization across scenes and cloth items.</p> <table> <tbody> <tr> <td>&nbsp;</td> <td># Scenes</td> <td>&nbsp;</td> <td># Clothes</td> <td>&nbsp;</td> <td># Images</td> <td>&nbsp;</td> </tr> <tr> <td>&nbsp;</td> <td>Train</td> <td>Test</td> <td>Train</td> <td>Test</td> <td>Train</td> <td>Test</td> </tr> <tr> <td> <p>Tshirts</p> </td> <td>6</td> <td>8</td> <td>15</td> <td>20</td> <td>210</td> <td>400</td> </tr> <tr> <td>Towels</td> <td>6</td> <td>8</td> <td>15</td> <td>20</td> <td>210</td> <td>400</td> </tr> <tr> <td>Shorts</td> <td>6</td> <td>8</td> <td>8</td> <td>9</td> <td>112</td> <td>180</td> </tr> <tr> <td>Boxershorts</td> <td>6</td> <td>8</td> <td>11</td> <td>11</td> <td>154</td> <td>220</td> </tr> <tr> <td>Total</td> <td>6</td> <td>8</td> <td>49</td> <td>60</td> <td>686</td> <td>1200</td> </tr> </tbody> </table> <p>&nbsp;</p> <p>* Not all boxershort images have been labeled yet.</p> <h3>Codebase&nbsp;</h3> <p>The codebased used to capture the images and label the dataset is available&nbsp;<a href="https://github.com/tlpss/aRTF-Clothes-dataset" target="_blank" rel="noopener">here</a>.&nbsp;&nbsp;</p> <h3>Files</h3> <p>There are three .zip files available:</p> <ul> <li>`aRTFClothes-rgb.zip` contains all full-sized images, with COCO and CVAT annotations.</li> <li>&nbsp;`aRTFClothes-resized-paper-splits` contains 3 zips that each contain the train/val/test splits (as coco datasets) that were used in the paper. All images are resized to 512x256.</li> <li>`artf-all-categories_resized_512x256.zip` contains train/val/test splits that combine the cloth categories</li> </ul> <p>&nbsp;</p> <p>The dataset was released in <a href="https://arxiv.org/abs/2401.01734">this</a> paper, if you use the dataset in scientific publications we ask to cite this work.</p>