{"id":31800,"date":"2020-04-25T16:55:36","date_gmt":"2020-04-25T08:55:36","guid":{"rendered":"\/blog\/?p=31800"},"modified":"2020-04-25T16:57:52","modified_gmt":"2020-04-25T08:57:52","slug":"axelerate-keras-based-framework-for-ai-on-the-edge","status":"publish","type":"post","link":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/","title":{"rendered":"aXeleRate \u2013 Keras-Based Framework for AI on the Edge"},"content":{"rendered":"\n<p><em>Train your machine learning models in Google Colab and easily optimize them for hardware accelerated inference!<\/em><\/p>\n\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe title=\"aXeleRate - Keras-Based Framework for AI on the Edge\" width=\"640\" height=\"360\" src=\"https:\/\/www.youtube.com\/embed\/8lp_45duWmM?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<p><em>Original article by Dmitry Maslov published on<\/em>&nbsp;<a href=\"https:\/\/www.hackster.io\/dmitrywat\/the-road-from-line-following-to-lane-following-de8836\" target=\"_blank\" rel=\"noreferrer noopener\">Hackster.io<\/a><\/p>\n\n\n\n<p>These might be very difficult times for many of us, depending on the part of the world you live in. Due to aggravating coronovirus pandemic many countries implemented strict lockdown policies. I myself recently had to spend 14 days in quarantine, staying indoors for 24 hours a day. I decided to make most of it and continue working on the stuff I am excited about, i.e. robotics and machine learning. And this is how&nbsp;<a href=\"https:\/\/github.com\/AIWintermuteAI\/aXeleRate\">aXeleRate<\/a>&nbsp;was born.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/hackster.imgix.net\/uploads\/attachments\/1095652\/capture_jfmnpdm5d6.JPG?auto=compress%2Cformat&amp;w=740&amp;h=555&amp;fit=max\" alt=\"\" \/><\/figure>\n\n\n\n<p>aXeleRate started as a personal project of mine for training YOLOv2 based object detection networks and exporting them to.kmodel format to be run on K210 chip. I also needed to train image classification networks. And sometimes I needed to run inference with Tensorflow Lite on Raspberry Pi. As a result I had a whole bunch of disconnected scripts with somewhat overlapping functionality. So, I decided to fix that by combining all the elements in an easy to use package and as a bonus part \u2013 making it fully compatible with Google Colab.<\/p>\n\n\n\n<p>aXeleRate is meant for people who need to run computer vision applications (image classification, object detection, semantic segmentation) on the edge devices with hardware acceleration. It has easy configuration process through config file or config dictionary (for Google Colab) and automatic conversion of the best model for training session into the required file format. You put the properly formatted data in, start the training script and come back to see a converted model that is ready for deployment on your device!<\/p>\n\n\n\n<p>Here is quick rundown of the features:<\/p>\n\n\n\n<p><strong>Key Features<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>Supports multiple computer vision models: object detection(YOLOv2), image classification, semantic segmentation(SegNet-basic)<\/li><li>Different feature extractors to be used with the above network types: Full Yolo, Tiny Yolo, MobileNet, SqueezeNet, VGG16, ResNet50, and Inception3.<\/li><li>Automatic conversion of the best model for the training session. aXeleRate will download the suitable converter automatically.<\/li><li>Currently supports trained model conversion to:.kmodel(K210),.tflite formats. Support planned for:.tflite(Edge TPU),.pb(TF-TRT optimized).<\/li><li>Model version control made easier. Keras model files and converted models are saved in the project folder, grouped by the training date. Training history is saved as.png graph in the model folder.<\/li><li>Two modes of operation: locally, with train.py script and.json config file and remote, tailored for Google Colab, with module import and dictionary config.<\/li><\/ul>\n\n\n\n<p>In this article we\u2019re going to train a person detection model for use with K210 chip on cyberEye board installed on M.A.R.K. mobile platform. M.A.R.K. (I\u2019ll call it MARK in text) stands for&nbsp;<a href=\"https:\/\/www.tinkergen.com\/mark\">Make a Robot Kit and it is an educational robot platform<\/a>&nbsp;in development by TinkerGen education. I take part in the development of MARK and we\u2019re currently preparing to launch a Kickstarter campaign. One of the main features of MARK is making machine learning concepts and workflow more transparent and easier to understand and use for teachers and students.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/hackster.imgix.net\/uploads\/attachments\/1095765\/dreamy_blue_red_mark(1)_yd5Y59SEby.jpg?auto=compress%2Cformat&amp;w=740&amp;h=555&amp;fit=max\" alt=\"\" \/><\/figure>\n\n\n\n<p>As it was mentioned before, aXeleRate can be run on local computer or in Google Colab. We\u2019ll opt for running on Google Colab, since it simplifies the preparation step.<\/p>\n\n\n\n<p>Let\u2019s open the&nbsp;<a href=\"https:\/\/colab.research.google.com\/github\/AIWintermuteAI\/aXeleRate\/blob\/master\/resources\/aXeleRate_test_detector.ipynb\">sample notebook<\/a><\/p>\n\n\n\n<p>Go through the cells one by one to get the understanding of the workflow. This example trains a detection network on a tiny dataset that is included with aXeleRate. For our next step we need a bigger dataset to actually train a useful model.<\/p>\n\n\n\n<p>Open the&nbsp;<a href=\"https:\/\/colab.research.google.com\/drive\/1w9QtBAgpJrbbg1Sua9I2NAJIl34CXQHC\">notebook I prepared<\/a>. Follow the steps there and in the end, after a few hours of training you will get.h5,.tflite and.kmodel files saved in your Google Drive. Download the.kmodel file and copy it to an SD card and insert the SD card into mainboard. In our case with M.A.R.K. it is a modified version of Maixduino called cyberEye.<\/p>\n\n\n\n<p>MARK is an educational robot for introducing students to concepts of AI and machine learning. So, there are two ways to run a custom model you created just now: using Micropython code or our TinkerGen\u2019s graphical programming environment, called Codecraft. While the first one is undoubtedly more flexible in ways you can tweak the inference parameters, the second is more user-friendly.<\/p>\n\n\n\n<p>If you opt for graphical programming environment, then go to Codecraft website,&nbsp;<a href=\"https:\/\/ide.tinkergen.com\/\">https:\/\/ide.tinkergen.com&nbsp;<\/a>and choose MARK(cyberEye) as target platform.<\/p>\n\n\n\n<p>Click on Add Extension and choose Custom models, then click on Object Detection model.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/hackster.imgix.net\/uploads\/attachments\/1095677\/mark_withcustom_models_UZOGB7SgAp.png?auto=compress%2Cformat&amp;w=740&amp;h=555&amp;fit=max\" alt=\"\" \/><\/figure>\n\n\n\n<p>There you will need to enter filename of the model on SD card, the actual name of the model you will see in Codecraft interface(can be anything, let\u2019s enter Person detection), category name(person) and anchors. We didn\u2019t change anchors parameters, so will just use the default ones.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/hackster.imgix.net\/uploads\/attachments\/1095678\/object_detection_model_ul9rCcIMv5.png?auto=compress%2Cformat&amp;w=740&amp;h=555&amp;fit=max\" alt=\"\" \/><\/figure>\n\n\n\n<p>After that you will see three new blocks have appeared. Let\u2019s choose the one that outputs X coordinate of detected object and connect it to Display\u2026 at row 1. Put that inside of the loop and upload the code to MARK. You should see X coordinate of the center of the bounding box around detected person at the first row of the screen. If nothing is detected it will show -1.<\/p>\n\n\n\n<p>This allowed us to implement model inference in graphical programming environment. Now we\u2019ll go to Micropython and implement more advanced solution. Download and install MaixPy IDE from&nbsp;<a href=\"https:\/\/cn.dl.sipeed.com\/MAIX\/MaixPy\/ide\">here<\/a>.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/hackster.imgix.net\/uploads\/attachments\/1095681\/capture1_LaeqJmNfYg.JPG?auto=compress%2Cformat&amp;w=740&amp;h=555&amp;fit=max\" alt=\"\" \/><\/figure>\n\n\n\n<p>Open the example code I enclose with the article. The code logic is following:<\/p>\n\n\n\n<p>1) We check if there are people detected in find_center() function. If there are people found, it returns the x-coordinate of the center of the biggest detected bounding box. If no people detected, the function would return -1.<\/p>\n\n\n\n<p>2) If find_center() function returns the x-coordinate, we check if it is closer to image center\/on the left\/on the right, then control the motors accordingly.<\/p>\n\n\n\n<p>3) If find_center() function returns -1, we use servo to do 40 degree tilt scan for people with the camera.<\/p>\n\n\n\n<p>4) If while tilt scan we are unable to find people, the robot does 2 180 degree pan scans.<\/p>\n\n\n\n<p>5) Finally if pan scan doesn\u2019t detect any people, robot starts rotating in place in clockwise direction, while still performing person detection.<\/p>\n\n\n\n<p>Here is the final result of Micropython code in action! It could be improved still to make faster\/more robust.<\/p>\n\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-4-3 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe title=\"M.A.R.K. (K210) Person Following Example - Custom Model Inference\" width=\"640\" height=\"480\" src=\"https:\/\/www.youtube.com\/embed\/xnYg9Z0nNaE?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<p>aXeleRate is still work in progress project. I will be making some changes from time to time and if you find it useful and can contribute, PRs are very much welcome! In near future I will be making another video about inference on Raspberry Pi 4 with\/without hardware acceleration. We will have another WIP hardware appearance for that one. Which one? Can\u2019t tell you, hush-hush!<\/p>\n\n\n\n<p>Stay tuned for more articles from me and updates on MARK Kickstarter campaign.<\/p>\n\n\n\n<p><strong>For more information on Grove Zero series, Codecraft and other hardware for makers and STEM educators, visit our website,&nbsp;<a href=\"https:\/\/tinkergen.com\/\">https:\/\/tinkergen.com\/.<\/a><\/strong><\/p>\n\n\n\n<p>Until the next time and stay safe from the coronavirus!<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Code section<\/h2>\n\n\n\n<pre class=\"wp-block-code\"><code>import sensor,image,lcd, os, time, utime\nfrom maix_motor import Maix_motor\nimport KPU as kpu\n\ndef find_center():\n    img = sensor.snapshot()\n    img = img.resize(224,224)\n    img= img.rotation_corr(z_rotation=90.0)\n    a = img.pix_to_ai()\n    code = kpu.run_yolo2(task, img)\n    if code:\n        for i in code:\n            a=img.draw_rectangle(i.rect(),color = (0, 255, 0))\n            a = img.draw_string(i.x(),i.y(), classes[i.classid()], color=(255,0,0), scale=3)\n            x_center = i.x()+i.w()\/2\n            print(x_center)\n        a = lcd.display(img)\n        return x_center\n    else:\n        a = lcd.display(img)\n        return -1\n\n\ndef scan_pan():\n    Maix_motor.motor_run(0, 0, 0)\n    Maix_motor.servo_angle(2, 110)\n    for i in range(0, 180, 5):\n        pos = find_center()\n        if pos &gt;= 100 and pos &lt;= 124:\n            Maix_motor.servo_angle(1, 90)\n            if i &gt;= 90:\n                Maix_motor.motor_right(40, 0)\n                Maix_motor.motor_left(-40, 0)\n                utime.sleep_ms(((i - 90) * 12))\n            if i &lt; 90:\n                Maix_motor.motor_right(-40, 0)\n                Maix_motor.motor_left(40, 0)\n                utime.sleep_ms(((90 - i) * 12))\n            Maix_motor.motor_motion(1, 1, 0)\n            time.sleep(0.3)\n            return True\n        else:\n            Maix_motor.servo_angle(1, i)\n\n\ndef scan_tilt():\n    #Maix_motor.motor_run(0, 0, 0)\n    Maix_motor.servo_angle(1, 90)\n    for i in range(100, 140, 2):\n        pos = find_center()\n        if pos &gt; 0:\n            return True\n        else:\n            Maix_motor.servo_angle(2, i)\n\ndef scan_rotate():\n    Maix_motor.servo_angle(1, 90)\n    Maix_motor.servo_angle(2, 110)\n    Maix_motor.motor_motion(1, 3, 0)\n    while True:\n        pos = find_center()\n        if pos &gt; 0:\n            break\n\nlcd.init()\nsensor.reset()\nsensor.set_pixformat(sensor.RGB565)\nsensor.set_framesize(sensor.QVGA)\nsensor.set_vflip(1)\nsensor.run(1)\nMaix_motor.servo_angle(1, 90)\nMaix_motor.servo_angle(2, 110)\nDEBUG = 0\nclasses = [\"person\"]\ntask = kpu.load('\/sd\/model_people_end.kmodel')\nanchor = (0.57273, 0.677385, 1.87446, 2.06253, 3.33843, 5.47434, 7.88282, 3.52778, 9.77052, 9.16828)\na = kpu.init_yolo2(task, 0.3, 0.3, 5, anchor)\nwhile(True):\n    x_center = find_center()\n    if not DEBUG:\n        if x_center &gt;= 100 and x_center &lt;= 124: Maix_motor.motor_motion(1, 1, 0)\n        if x_center &lt; 100 and x_center &gt; 0: Maix_motor.motor_motion(1, 4, 0)\n        if x_center &gt; 124: Maix_motor.motor_motion(1, 3, 0)\n        if x_center &lt; 0:\n            if not scan_tilt():\n                scan_pan()\n                if not scan_pan():\n                    scan_rotate()\na = kpu.deinit(task)<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"<p>Train your machine learning models in Google Colab and easily optimize them for hardware accelerated<\/p>\n","protected":false},"author":3497,"featured_media":31801,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_lmt_disableupdate":"","_lmt_disable":"","_price":"","_stock":"","_tribe_ticket_header":"","_tribe_default_ticket_provider":"","_tribe_ticket_capacity":"0","_ticket_start_date":"","_ticket_end_date":"","_tribe_ticket_show_description":"","_tribe_ticket_show_not_going":false,"_tribe_ticket_use_global_stock":"","_tribe_ticket_global_stock_level":"","_global_stock_mode":"","_global_stock_cap":"","_tribe_rsvp_for_event":"","_tribe_ticket_going_count":"","_tribe_ticket_not_going_count":"","_tribe_tickets_list":"[]","_tribe_ticket_has_attendee_info_fields":false,"iawp_total_views":0,"footnotes":""},"categories":[1],"tags":[3047,3133,3095,3132,1608,2244,1086,795,826,3098],"class_list":["post-31800","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-codecraft","tag-codekids","tag-edtech","tag-m-a-r-k","tag-machine-learning","tag-machine-vision","tag-robotics","tag-steam","tag-stem-education","tag-tinkergen"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v24.0 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio\" \/>\n<meta property=\"og:description\" content=\"Train your machine learning models in Google Colab and easily optimize them for hardware accelerated\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/\" \/>\n<meta property=\"og:site_name\" content=\"Latest News from Seeed Studio\" \/>\n<meta property=\"article:published_time\" content=\"2020-04-25T08:55:36+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2020-04-25T08:57:52+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1280\" \/>\n\t<meta property=\"og:image:height\" content=\"400\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"leslie\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"leslie\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/\",\"url\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/\",\"name\":\"aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio\",\"isPartOf\":{\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png\",\"datePublished\":\"2020-04-25T08:55:36+00:00\",\"dateModified\":\"2020-04-25T08:57:52+00:00\",\"author\":{\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/773b99c9e7dd26504c50baac761c4400\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage\",\"url\":\"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png\",\"contentUrl\":\"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png\",\"width\":1280,\"height\":400},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.seeedstudio.com\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"aXeleRate \u2013 Keras-Based Framework for AI on the Edge\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/#website\",\"url\":\"https:\/\/www.seeedstudio.com\/blog\/\",\"name\":\"Latest News from Seeed Studio\",\"description\":\"Emerging IoT, AI and Autonomous Applications on the Edge\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.seeedstudio.com\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/773b99c9e7dd26504c50baac761c4400\",\"name\":\"leslie\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/668c09b44fc5039121c14f6dd35d0f10?s=96&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/668c09b44fc5039121c14f6dd35d0f10?s=96&r=g\",\"caption\":\"leslie\"},\"url\":\"https:\/\/www.seeedstudio.com\/blog\/author\/leslie\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/","og_locale":"en_US","og_type":"article","og_title":"aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio","og_description":"Train your machine learning models in Google Colab and easily optimize them for hardware accelerated","og_url":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/","og_site_name":"Latest News from Seeed Studio","article_published_time":"2020-04-25T08:55:36+00:00","article_modified_time":"2020-04-25T08:57:52+00:00","og_image":[{"width":1280,"height":400,"url":"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png","type":"image\/png"}],"author":"leslie","twitter_card":"summary_large_image","twitter_misc":{"Written by":"leslie","Est. reading time":"8 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/","url":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/","name":"aXeleRate \u2013 Keras-Based Framework for AI on the Edge - Latest News from Seeed Studio","isPartOf":{"@id":"https:\/\/www.seeedstudio.com\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage"},"image":{"@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage"},"thumbnailUrl":"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png","datePublished":"2020-04-25T08:55:36+00:00","dateModified":"2020-04-25T08:57:52+00:00","author":{"@id":"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/773b99c9e7dd26504c50baac761c4400"},"breadcrumb":{"@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#primaryimage","url":"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png","contentUrl":"https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png","width":1280,"height":400},{"@type":"BreadcrumbList","@id":"https:\/\/www.seeedstudio.com\/blog\/2020\/04\/25\/axelerate-keras-based-framework-for-ai-on-the-edge\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.seeedstudio.com\/blog\/"},{"@type":"ListItem","position":2,"name":"aXeleRate \u2013 Keras-Based Framework for AI on the Edge"}]},{"@type":"WebSite","@id":"https:\/\/www.seeedstudio.com\/blog\/#website","url":"https:\/\/www.seeedstudio.com\/blog\/","name":"Latest News from Seeed Studio","description":"Emerging IoT, AI and Autonomous Applications on the Edge","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.seeedstudio.com\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/773b99c9e7dd26504c50baac761c4400","name":"leslie","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.seeedstudio.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/668c09b44fc5039121c14f6dd35d0f10?s=96&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/668c09b44fc5039121c14f6dd35d0f10?s=96&r=g","caption":"leslie"},"url":"https:\/\/www.seeedstudio.com\/blog\/author\/leslie\/"}]}},"modified_by":"leslie","views":6621,"featured_image_urls":{"full":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png",1280,400,false],"thumbnail":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-80x80.png",80,80,true],"medium":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-300x94.png",300,94,true],"medium_large":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-768x240.png",640,200,true],"large":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-1030x322.png",640,200,true],"1536x1536":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png",1280,400,false],"2048x2048":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png",1280,400,false],"visody_icon":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png",32,10,false],"magazine-7-slider-full":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1.png",1280,400,false],"magazine-7-slider-center":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-936x400.png",936,400,true],"magazine-7-featured":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-1024x320.png",1024,320,true],"magazine-7-medium":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-720x380.png",720,380,true],"magazine-7-medium-square":["https:\/\/www.seeedstudio.com\/blog\/wp-content\/uploads\/2020\/04\/3-1-675x400.png",675,400,true]},"author_info":{"display_name":"leslie","author_link":"https:\/\/www.seeedstudio.com\/blog\/author\/leslie\/"},"category_info":"<a href=\"https:\/\/www.seeedstudio.com\/blog\/category\/news\/\" rel=\"category tag\">News<\/a>","tag_info":"News","comment_count":"0","_links":{"self":[{"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/posts\/31800","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/users\/3497"}],"replies":[{"embeddable":true,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/comments?post=31800"}],"version-history":[{"count":1,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/posts\/31800\/revisions"}],"predecessor-version":[{"id":31802,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/posts\/31800\/revisions\/31802"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/media\/31801"}],"wp:attachment":[{"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/media?parent=31800"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/categories?post=31800"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.seeedstudio.com\/blog\/wp-json\/wp\/v2\/tags?post=31800"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}