{"id":2530,"date":"2021-12-24T09:00:00","date_gmt":"2021-12-24T01:00:00","guid":{"rendered":"https:\/\/blog.ailia.ai\/uncategorized\/sku110k-densedet-machine-learning-model-that-can-detect-products-in-a-store\/"},"modified":"2025-05-20T21:18:48","modified_gmt":"2025-05-20T13:18:48","slug":"sku110k-densedet-machine-learning-model-that-can-detect-products-in-a-store","status":"publish","type":"post","link":"https:\/\/blog.ailia.ai\/en\/tips-en\/sku110k-densedet-machine-learning-model-that-can-detect-products-in-a-store\/","title":{"rendered":"SKU110K-DenseDet : A Machine Learning Model That Can Detect Products in a Store."},"content":{"rendered":"\n

Overview<\/strong><\/h3>\n\n\n\n

SKU110K-DenseDet<\/em>\u00a0is a machine learning model that can detect products in a store. It can detect the bounding boxes of products presented on a supermarket shelf, but there is no categorization, only the presence of a product is detected.<\/p>\n\n\n\n

\"\"
Source:\u00a0https:\/\/github.com\/Media-Smart\/SKU110K-DenseDet<\/a><\/figcaption><\/figure>\n\n\n\n

<\/a>A Solution to Product detection in Densely Packed Scenes<\/a><\/p>\n\n\n\n

<\/a>GitHub – Media-Smart\/SKU110K-DenseDet<\/a><\/p>\n\n\n\n

About SKU110K<\/strong><\/h3>\n\n\n\n

SKU110K\u00a0<\/em>is a data set for product detection published in April 2019 which contains 11,762 images taken with cell phones in thousands of supermarkets around the world (United States, Europe, East Asia). Bounding boxes were manually annotated. It contains 90,968 bounding boxes in 8,233 images for training and 432,312 bounding boxes in 2,941 images for validation.<\/p>\n\n\n\n

\"\"
Source:\u00a0https:\/\/arxiv.org\/pdf\/1904.00853.pdf<\/a><\/figcaption><\/figure>\n\n\n\n

<\/a>GitHub – eg4000\/SKU110K_CVPR19<\/a><\/p>\n\n\n\n

<\/a>Precise Detection in Densely Packed Scenes<\/a><\/p>\n\n\n\n

Architecture<\/strong><\/h3>\n\n\n\n

SKU110K-DenseDet <\/em>was trained using MMDetection<\/em>. It achieves a 58.0% mAP using Cascade R-CNN<\/a>. The backbone uses ResNXt-101<\/em>.<\/p>\n\n\n\n

Since SKU110K <\/em>contains many small objects, usual architectures and input resolutions for object detection is not accurate enough. Therefore input size of images is set to 2560×2560.<\/p>\n\n\n\n

In environments with low GPU memory, random cropping of images is used during training, in a way that it does not negatively impact the training results.<\/p>\n\n\n\n

SKU110K <\/em>contains on average 150 bounding boxes per image. This is significantly more than MS COCO<\/em> and default hyper parameters are not optimal. Hence the max positive sample number of both RPN and R-CNN sampler were adjusted to release the limits.<\/p>\n\n\n\n

Usage<\/strong><\/h3>\n\n\n\n

SKU110K-DenseDet\u00a0<\/em>can be executed using ailia SDK with the following command. Due to the huge size of the backbone, please specify\u00a0-e 0<\/code>\u00a0to run in CPU mode in environments with low GPU memory.<\/p>\n\n\n\n

$ python3 sku110k-densedet.py -i input.jpg -e 0<\/code><\/p>\n\n\n\n

<\/a>ailia-models\/object_detection\/sku110k-densedet<\/a><\/p>\n\n\n\n

ax Inc.<\/a> has developed ailia SDK<\/a>, which enables cross-platform, GPU-based rapid inference.<\/p>\n\n\n\n

ax Inc. provides a wide range of services from consulting and model creation, to the development of AI-based applications and SDKs. Feel free to contact us<\/a> for any inquiry.<\/p>\n","protected":false},"excerpt":{"rendered":"

Overview SKU110K-DenseDet\u00a0is a machine learning model that can detect products in a store. It can detect the bounding boxes of products presented on a supermarket shelf, but there is no categorization, only the presence of a product is detected. A Solution to Product detection in Densely Packed Scenes GitHub – Media-Smart\/SKU110K-DenseDet About SKU110K SKU110K\u00a0is a data set for product detection published in April 2019 which contains 11,762 images taken with cell phones in thousands of supermarkets around the world (United States, Europe, East Asia). Bounding boxes were manually annotated. It contains 90,968 bounding boxes in 8,233 images for training and 432,312 bounding boxes in 2,941 images for validation. GitHub – […]<\/p>\n","protected":false},"author":5,"featured_media":2444,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[255],"tags":[266],"class_list":["post-2530","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tips-en","tag-ailiamodels-en"],"acf":[],"_links":{"self":[{"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/posts\/2530","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/comments?post=2530"}],"version-history":[{"count":1,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/posts\/2530\/revisions"}],"predecessor-version":[{"id":2532,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/posts\/2530\/revisions\/2532"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/media\/2444"}],"wp:attachment":[{"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/media?parent=2530"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/categories?post=2530"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.ailia.ai\/en\/wp-json\/wp\/v2\/tags?post=2530"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}