{"id":359557,"date":"2022-03-08T12:00:00","date_gmt":"2022-03-08T11:00:00","guid":{"rendered":"https:\/\/innovationorigins.com\/?post_type=selected&p=359557"},"modified":"2022-03-08T12:00:00","modified_gmt":"2022-03-08T11:00:00","slug":"neural-networks-to-improve-local-image-recognition-on-phones","status":"publish","type":"selected","link":"https:\/\/ioplus.nl\/archive\/en\/selected\/neural-networks-to-improve-local-image-recognition-on-phones\/","title":{"rendered":"Neural networks to improve local image recognition on phones"},"content":{"rendered":"\n

We use our smartphones every day to scan a multitude of objects and access a wide range of services thanks to image recognition technology using artificial intelligence. But these applications require access to the mobile network to connect to the cloud and request the classification response from remote servers. Three companies (IMATAG, QUAI DES APPS and ARIADNEXT) have joined forces in the Inria Rennes \u2013 Bretagne Atlantique research center to solve this problem. They have optimized deep learning algorithms for execution on embedded platforms while still maintaining a good level of performance, as announced by the French National Institute for Research in Digital Science and Technology in a press release.<\/a><\/p>\n\n\n\n

From cloud to local<\/h3>\n\n\n\n

\u201cCurrently, in most approaches used to\u00a0classify images\u00a0or read identity documents from a smartphone, the device is only used to capture the image. The data is sent to cloud servers for analysis, a process that is often computationally intensive. The result is then sent back to the user\u2019s telephone. The technique works well so long as network coverage is available, but not in areas with no coverage or where access is limited. <\/p>\n\n\n\n

To solve this problem, the\u00a0MobileAI\u00a0research project aimed to incorporate\u00a0artificial intelligence\u00a0technology into the smartphone while maintaining its robustness and ability to operate in real time,\u201d explains Montaser Awal, head of the artificial intelligence research team at\u00a0ARIADNEXT, a company specializing in the remote verification of ID documents.<\/p>\n\n\n\n

\u201cThe project was born from informal discussions among a group of people working at three different companies where visual content identification plays a central role,\u201d\u00a0says Mathieu Desoubeaux, Co-Founder of\u00a0IMATAG<\/a>, a company created with the support of Inria that specializes in robust watermarks for copyright-protected content. \u201cThe subject was first mentioned by our friends at\u00a0QUAI DES APPS<\/a>, a company that works in the field of augmented reality. Their aim is to achieve image recognition on a mobile phone without network coverage. Yannis Avrithis, a researcher from the\u00a0Linkmedia<\/a>\u00a0team at the Inria Rennes Centre also played a very active role in these discussions. That was how the four of us decided to set up an R&D collaboration to try to solve the problem.\u201d<\/p>\n\n\n\n

Launched in September 2018 and completed in 2021, the\u00a0MobileAI\u00a0project was funded by the BPI, Rennes M\u00e9tropole and the regions of Brittany and Pays de la Loire through a call for projects launched by Images & R\u00e9seaux.<\/p>\n\n\n\n

Deep-learning algorithms<\/h3>\n\n\n\n

At the heart of the matter is a family of particularly powerful\u00a0deep-learning algorithms, called convolutional neural networks (CNN). \u201cThese are excellent candidates for mobile image recognition\u201d,\u00a0explains Montaser Awal. \u201cBut for our purposes, we had to modify their architecture and optimize them to make them executable on mobile devices while maintaining a similar performance level to\u00a0cloud-based server systems.\u201d<\/p>\n\n\n\n

Mission accomplished. The project has advanced the state of the art and resulted in ten scientific publications and five prototype applications. The new algorithms for image classification and text recognition from a photograph of an ID document were immediately integrated into\u00a0IDcheck.io<\/a><\/em>,\u00a0ARIADNEXT’s flagship product for ID document authentication. \u201cThe acquirement of cutting-edge expertise in deep learning for image recognition is also an important factor for future developments,\u201d the company explains.<\/p>\n\n\n\n

The project has allowed QUAI DES APPS to improve\u00a0Blinkl<\/a>, its augmented narration web app. Its service allows clients in shops to photograph products on the shelves and obtain more information about them. Until now, the image recognition process was executed on remote servers. The disadvantages of this were the computational load on these machines and latencies during peak periods, such as during sales or product launches. In addition, there was a bottleneck in the image search which limited the size of the database to 1000 products. By switching image recognition to mobile and improving the descriptors of these images, the company made a game-changing move and can now handle databases of 100,000 images. These capacities will allow QUAI DES APPS to meet the needs of the retail industry, whether for images of products on the shelf or in a catalogue.<\/p>\n\n\n\n

For IMATAG, this R&D project improved the image search technology used in its\u00a0monitoring solution for copyright infringement<\/a>. It also opens up prospects for new product lines.<\/p>\n\n\n\n

Also interesting:<\/p>\n\n\n\n

Making mechanical ventilators smarter through data-driven algorithms<\/a>
The disease-predicting power of algorithms<\/a>
Machine-learning methods help Science better understand solar panels<\/a><\/p>\n\n\n\n

<\/p>\n","protected":false},"author":1730,"featured_media":359561,"template":"","meta":{"_acf_changed":false,"advgb_blocks_editor_width":"","advgb_blocks_columns_visual_guide":""},"categories":[8553,26049],"tags":[28622,28724,49457,28726],"location":[40146],"internal_archives":[],"class_list":["post-359557","selected","type-selected","status-publish","has-post-thumbnail","hentry","category-digital","category-mobility","tag-deep-learning-en","tag-image-recognition","tag-mobile-phones","tag-neural-networks","location-france"],"blocksy_meta":[],"acf":[],"featured_img":false,"coauthors":[],"author_meta":{"author_link":"https:\/\/ioplus.nl\/archive\/author\/brenda-arnold\/","display_name":"Brenda Arnold"},"relative_dates":{"created":"Posted 4 years ago","modified":"Updated 4 years ago"},"absolute_dates":{"created":"Posted on March 8, 2022","modified":"Updated on March 8, 2022"},"absolute_dates_time":{"created":"Posted on March 8, 2022 12:00 pm","modified":"Updated on March 8, 2022 12:00 pm"},"featured_img_caption":"","tax_additional":{"category":{"linked":["Digital<\/a>","Mobility<\/a>"],"unlinked":["Digital<\/span>","Mobility<\/span>"],"slug":"category","name":"Categories"},"post_tag":{"linked":["deep learning<\/a>","image recognition<\/a>","mobile phones<\/a>","neural networks<\/a>"],"unlinked":["deep learning<\/span>","image recognition<\/span>","mobile phones<\/span>","neural networks<\/span>"],"slug":"post_tag","name":"Tags"},"language":{"linked":["EN<\/a>"],"unlinked":["EN<\/span>"],"slug":"language","name":"Tags"},"post_translations":{"linked":["pll_622728a233e5b<\/a>"],"unlinked":["pll_622728a233e5b<\/span>"],"slug":"post_translations","name":""},"location":{"linked":["France<\/a>"],"unlinked":["France<\/span>"],"slug":"location","name":"Locations"},"internal_archives":{"linked":[],"unlinked":[],"slug":"internal_archives","name":"Internal Archives"}},"series_order":"","_links":{"self":[{"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/selected\/359557","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/selected"}],"about":[{"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/types\/selected"}],"author":[{"embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/users\/1730"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/"}],"wp:attachment":[{"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/media?parent=359557"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/categories?post=359557"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/tags?post=359557"},{"taxonomy":"location","embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/location?post=359557"},{"taxonomy":"internal_archives","embeddable":true,"href":"https:\/\/ioplus.nl\/archive\/wp-json\/wp\/v2\/internal_archives?post=359557"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}