ImageNet16: Small scale ImageNet Classification

<p>This is a subset of ImageNet called "ImageNet16" more suited for cases with limited computational budget and faster experimentation.&nbsp;</p> <p>Each class has 400 train images and 100 test images.</p> <p>* Credit also goes to original creators that constructed the dataset. Unfortunately, I was not able to relocated it online so I reupload it here.</p> <p>&nbsp;</p> <p>If used in your work please cite as follows:</p> <p>C. Kyrkou, "Toward Efficient Convolutional Neural Networks With Structured Ternary Patterns," in&nbsp;<em>IEEE Transactions on Neural Networks and Learning Systems</em>, doi: 10.1109/TNNLS.2024.3380827.</p> <p>&nbsp;</p> <p>The classes corresponding to imagenet1K:</p> <p>&bull; n02009912 American_egret</p> <p>&bull;&nbsp;n02113624 toy_poodle</p> <p>&bull;&nbsp;&nbsp;n02123597 Siamese_cat</p> <p>&bull;&nbsp;n02132136 brown_bear</p> <p>&bull;&nbsp;n02504458 African_elephant</p> <p>&bull;&nbsp;n02690373 airliner</p> <p>&bull;&nbsp;n02835271 bicycle-built-for-two</p> <p>&bull;&nbsp;n02951358 canoe</p> <p>&bull;&nbsp;n03041632 cleaver</p> <p>&bull;&nbsp;n03085013 computer_keyboard</p> <p>&bull;&nbsp;n03196217 digital_clock</p> <p>&bull;&nbsp;n03977966 police_van</p> <p>&bull;&nbsp;n04099969 rocking_chair</p> <p>&bull;&nbsp;n04111531 rotisserie</p> <p>&bull;&nbsp;n04285008 sports_car</p> <p>&bull;&nbsp;n04591713 wine_bottle</p> <p>&nbsp;</p> <p><strong>From original map.txt</strong></p> <p>knife =&nbsp;&nbsp;&nbsp; n03041632</p> <p>keyboard = n03085013</p> <p>elephant = n02504458</p> <p>bicycle =&nbsp; n02835271</p> <p>airplane = n02690373</p> <p>clock =&nbsp;&nbsp;&nbsp; n03196217</p> <p>oven =&nbsp;&nbsp;&nbsp;&nbsp; n04111531</p> <p>chair =&nbsp;&nbsp;&nbsp; n04099969</p> <p>bear =&nbsp;&nbsp;&nbsp;&nbsp; n02132136</p> <p>boat =&nbsp;&nbsp;&nbsp;&nbsp; n02951358</p> <p>cat =&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; n02123597</p> <p>bottle =&nbsp;&nbsp; n04591713</p> <p>truck =&nbsp;&nbsp;&nbsp; n03977966</p> <p>car =&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; n04285008</p> <p>bird =&nbsp;&nbsp;&nbsp;&nbsp; n02009912</p> <p>dog =&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; n02113624</p> <p>&nbsp;</p> <p><strong>Folder Structure</strong></p> <p>-&lt;train&gt;</p> <p>&nbsp; -- &lt;class idx 1&gt;</p> <p>&nbsp; &nbsp; &nbsp; --- &lt;filename1&gt;.JPEG</p> <p>&nbsp; &nbsp; &nbsp; --- &lt;filename2&gt;.JPEG</p> <p>&nbsp; &nbsp; &nbsp; --- ....</p> <p>&nbsp; -- &lt;class idx 1&gt;</p> <p>&nbsp; --...</p> <p>&nbsp;</p> <p>-&lt;val&gt;</p> <p>&nbsp; -- &lt;class idx 1&gt;</p> <p>&nbsp; &nbsp; &nbsp; --- &lt;filename1&gt;.JPEG</p> <p>&nbsp; &nbsp; &nbsp; --- &lt;filename2&gt;.JPEG</p> <p>&nbsp; &nbsp; &nbsp; --- ....</p> <p>&nbsp; -- &lt;class idx 1&gt;</p> <p>&nbsp; --...</p> <p>&nbsp;</p> <p><strong>Some preliminary results:</strong></p> <table> <tbody> <tr> <td><strong>Model Name</strong></td> <td><strong>Accuracy (Top-1)</strong></td> </tr> <tr> <td>VGG16</td> <td>85.3</td> </tr> <tr> <td>ResNet50</td> <td>88.2</td> </tr> <tr> <td>MobileNetV2</td> <td>91.0</td> </tr> <tr> <td>EfficientNet B0</td> <td>85.6</td> </tr> </tbody> </table> <p>&nbsp;</p> <p><em>Massive Credit to original ImageNet authors</em><br>[1]&nbsp;Olga Russakovsky<em>, Jia Deng</em>, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, Alexander C. Berg and Li Fei-Fei.ImageNet Large Scale Visual Recognition Challenge. IJCV, 2015</p>