{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/colornet-investigating-the-importance-of","title":"ColorNet: Investigating the importance of color spaces for image classification","arxiv_id":"1902.00267","date":"2019-02-01","proceeding":null,"authors":["Shreyank N Gowda","Chun Yuan"],"abstract":"Image classification is a fundamental application in computer vision.\nRecently, deeper networks and highly connected networks have shown state of the\nart performance for image classification tasks. Most datasets these days\nconsist of a finite number of color images. These color images are taken as\ninput in the form of RGB images and classification is done without modifying\nthem. We explore the importance of color spaces and show that color spaces\n(essentially transformations of original RGB images) can significantly affect\nclassification accuracy. Further, we show that certain classes of images are\nbetter represented in particular color spaces and for a dataset with a highly\nvarying number of classes such as CIFAR and Imagenet, using a model that\nconsiders multiple color spaces within the same model gives excellent levels of\naccuracy. Also, we show that such a model, where the input is preprocessed into\nmultiple color spaces simultaneously, needs far fewer parameters to obtain high\naccuracy for classification. For example, our model with 1.75M parameters\nsignificantly outperforms DenseNet 100-12 that has 12M parameters and gives\nresults comparable to Densenet-BC-190-40 that has 25.6M parameters for\nclassification of four competitive image classification datasets namely:\nCIFAR-10, CIFAR-100, SVHN and Imagenet. Our model essentially takes an RGB\nimage as input, simultaneously converts the image into 7 different color spaces\nand uses these as inputs to individual densenets. We use small and wide\ndensenets to reduce computation overhead and number of hyperparameters\nrequired. We obtain significant improvement on current state of the art results\non these datasets as well.","url_abs":"http://arxiv.org/abs/1902.00267v1","url_pdf":"http://arxiv.org/pdf/1902.00267v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"colornet-investigating-the-importance-of","repo_url":"https://github.com/kini5gowda/ColorNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"concatenated-skip-connection","method_name":"Concatenated Skip Connection"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-block","method_name":"Dense Block"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-cifar-100","task":"Image Classification","dataset":"CIFAR-100","model":"ColorNet","rank_in_archive_order":37,"of":211,"metrics":{"PARAMS":"19.0M","Percentage correct":"88.4"},"uses_additional_data":true},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ColorNet (RHYLH with Conv Layer)","rank_in_archive_order":323,"of":1060,"metrics":{"Top 1 Accuracy":"84.32%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ColorNet","rank_in_archive_order":546,"of":1060,"metrics":{"Top 1 Accuracy":"82.35%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-svhn","task":"Image Classification","dataset":"SVHN","model":"Colornet","rank_in_archive_order":3,"of":62,"metrics":{"Percentage error":"1.11"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1902.00267","atlas_url":"https://app.syntology.ai/?focus=1902.00267","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}