{"id":164614,"date":"2022-06-01T09:54:00","date_gmt":"2022-06-01T08:54:00","guid":{"rendered":"https:\/\/datascientest.com\/?p=164549"},"modified":"2023-11-09T14:04:06","modified_gmt":"2023-11-09T14:04:06","slug":"mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","status":"publish","type":"post","link":"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","title":{"rendered":"Mushroom recognition : How computer vision can aid in the identification of species"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"164614\" class=\"elementor elementor-164614\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-3829a44 e-flex e-con-boxed e-con\" data-id=\"3829a44\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-5a13e87 elementor-widget elementor-widget-heading\" data-id=\"5a13e87\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h3 class=\"elementor-heading-title elementor-size-default\">Context<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7c8749f elementor-widget elementor-widget-text-editor\" data-id=\"7c8749f\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p>One of the most widely used and oldest species identification techniques is <strong>\u00ab\u00a0morphological identification\u00a0\u00bb<\/strong>, which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it.<\/p><p>An alternative to this identification technique could be computer vision algorithms. These algorithms allow us to identify<strong> the characteristics of an image in a fast and accurate way<\/strong>, so the identification of a species can be made only through a photo, making it more portable.<\/p><p>In the following, we try to identify different mushroom images with the help of computer vision. This work was realized together with my promotion colleague Cyril Vandenberghe and supervised by Adrien Ruimy. It corresponds to the <a href=\"https:\/\/datascientest.com\/en\/data-scientist-course\">DataScientest data scientist course<\/a> final project. The development took a hundred hours and allowed us to utilize our newly acquired knowledge in a practical application.<\/p><p>Project Authors : <a href=\"https:\/\/www.linkedin.com\/in\/karina-castillo-perez\/\">Karina Castillo<\/a> &amp; <a href=\"https:\/\/www.linkedin.com\/in\/vdbcyril\/\">Cyril Vandenberghe<\/a>.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-a105550 e-flex e-con-boxed e-con\" data-id=\"a105550\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-004a847 elementor-widget elementor-widget-heading\" data-id=\"004a847\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">The project<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bef64e9 elementor-widget elementor-widget-heading\" data-id=\"bef64e9\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Data source<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4357549 elementor-widget elementor-widget-text-editor\" data-id=\"4357549\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>For this project, we used data from <a href=\"https:\/\/mushroomobserver.org\/\">mushroomobserver.org<\/a>, a website dedicated to collect and identify mushroom images: Users upload images, and the community identifies them. Currently, it counts over 10,000 registered users.<\/p><p>During our initial bibliographical research, we found that a previous project hosted on Github, <a href=\"https:\/\/github.com\/bechtle\/mushroomobser-dataset\">\u00ab mushroomobser-dataset \u00bb<\/a> had already explored this site to create a dataset of images dating from 2006 to 2016. Additionally, 12 JSON files contain additional information about each image.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b17e163 elementor-widget elementor-widget-image\" data-id=\"b17e163\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"608\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-v2pJ4VymwVI-F55xa1b2rA-1024x778.png\" class=\"attachment-large size-large wp-image-177917\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-v2pJ4VymwVI-F55xa1b2rA-1024x778.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-v2pJ4VymwVI-F55xa1b2rA-300x228.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-v2pJ4VymwVI-F55xa1b2rA-768x584.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-v2pJ4VymwVI-F55xa1b2rA.png 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 1. Example of a JSON entry for one image. (image by Github: bechtle).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c337a36 elementor-widget elementor-widget-heading\" data-id=\"c337a36\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Data exploration<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-ef5f3a0 elementor-widget elementor-widget-text-editor\" data-id=\"ef5f3a0\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The first step in processing our data was to create a data frame with information about the images and <strong>taxonomy<\/strong> of the mushrooms. In biology, a taxonomy is a way of defining hierarchical groups based on common characteristics, as shown in the figure below :<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9186859 elementor-widget elementor-widget-image\" data-id=\"9186859\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"399\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-QAW4Cc-KLRAr6rIb0Yix7w-1024x511.jpeg\" class=\"attachment-large size-large wp-image-177918\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-QAW4Cc-KLRAr6rIb0Yix7w-1024x511.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-QAW4Cc-KLRAr6rIb0Yix7w-300x150.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-QAW4Cc-KLRAr6rIb0Yix7w-768x383.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-QAW4Cc-KLRAr6rIb0Yix7w.jpeg 1255w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 2. Taxonomy organization (Image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6f61294 elementor-widget elementor-widget-text-editor\" data-id=\"6f61294\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>In the table 1, the <em>`rank`<\/em> column indicates the last taxonomic level reached in the mushroom identification in our data frame. Thus, if an entry has the value <strong><em>`KINGDOM`<\/em><\/strong> in the <em>`rank`<\/em> column, the <em>`label`<\/em> column value will be the most general identification, e.g., <em>`Fungi.`<\/em> The identified images can therefore belong to any mushroom.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b675198 elementor-widget elementor-widget-image\" data-id=\"b675198\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"232\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-gy4Ol0Ho6HvQ6A4j0zYYNA-1024x297.jpeg\" class=\"attachment-large size-large wp-image-177920\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-gy4Ol0Ho6HvQ6A4j0zYYNA-1024x297.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-gy4Ol0Ho6HvQ6A4j0zYYNA-300x87.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-gy4Ol0Ho6HvQ6A4j0zYYNA-768x223.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-gy4Ol0Ho6HvQ6A4j0zYYNA.jpeg 1260w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Table 1. Example of values in the data frame when \"rank\" = \"KINGDOM\".<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-81d7a7f elementor-widget elementor-widget-text-editor\" data-id=\"81d7a7f\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>In contrast, if the <em>`rank`<\/em> column is <strong><em>`SPECIES`<\/em> <\/strong>(a more specific taxon), we have a more detailed description of the images (table 2).<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4b8e503 elementor-widget elementor-widget-image\" data-id=\"4b8e503\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"159\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-1c5S6Y_YwSZhVHCBHUhR4g-1024x204.jpeg\" class=\"attachment-large size-large wp-image-177921\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-1c5S6Y_YwSZhVHCBHUhR4g-1024x204.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-1c5S6Y_YwSZhVHCBHUhR4g-300x60.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-1c5S6Y_YwSZhVHCBHUhR4g-768x153.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-1c5S6Y_YwSZhVHCBHUhR4g.jpeg 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Table 2. Example of values in the data frame when \"rank\" = \"SPECIES\".<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d195303 elementor-widget elementor-widget-text-editor\" data-id=\"d195303\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Therefore, images with <em><strong>`KINGDOM`<\/strong><\/em>, <em><strong>`PHYLUM`<\/strong><\/em>, <em><strong>`CLASS`<\/strong> <\/em>and <em><strong>`ORDER`<\/strong> <\/em>values in the <em>`rank`<\/em> column were discarded, as they are too general and non-informative. For the rest of the project, we will only work with data belonging to the taxa <em>SPECIES, GENUS,<\/em> and <em>FAMILY<\/em> as shown in the figure below :<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-938500c elementor-widget elementor-widget-image\" data-id=\"938500c\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"399\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ifuPCIhGO7v68XxXopolUA-1024x511.jpeg\" class=\"attachment-large size-large wp-image-177922\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ifuPCIhGO7v68XxXopolUA-1024x511.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ifuPCIhGO7v68XxXopolUA-300x150.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ifuPCIhGO7v68XxXopolUA-768x383.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ifuPCIhGO7v68XxXopolUA.jpeg 1254w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 3. Selected taxa (Image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2c12ef8 elementor-widget elementor-widget-heading\" data-id=\"2c12ef8\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Data Visualisation<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4aa57cd elementor-widget elementor-widget-text-editor\" data-id=\"4aa57cd\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Having a large number of images is essential for successfully developing a <strong>computer vision<\/strong> project. In this project, we need labels (i.e., values in the <em>`label`<\/em> column) to have a large number of images to represent it for each taxon.<\/p><p>The image below (fig. 4) is a visualization of the distribution of the number of images in the taxa. Each of the four panels shows the number of labels<strong> (y-axis)<\/strong> for a given taxon <strong>(x-axis)<\/strong> represented by 1 to 10 images, 11 to 100 images, 101 to 1000 images, and over 1000 images.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-74e8e5d elementor-widget elementor-widget-image\" data-id=\"74e8e5d\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"250\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-i8SP6yWXRGAW2J1kivj_mg-1024x320.png\" class=\"attachment-large size-large wp-image-177923\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-i8SP6yWXRGAW2J1kivj_mg-1024x320.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-i8SP6yWXRGAW2J1kivj_mg-300x94.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-i8SP6yWXRGAW2J1kivj_mg-768x240.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-i8SP6yWXRGAW2J1kivj_mg.png 1382w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">FIg. 4. Number of labels per taxon by range. (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7bd6bae elementor-widget elementor-widget-text-editor\" data-id=\"7bd6bae\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The species taxon is the ideal one because it is the most specific one. However, the number of labels represented by only <strong>1 to 10 images corresponds to 75% of the total number of species<\/strong>. In the interval [101\u20131000], only seven species have more than 500 images, and only one species has more than 1000 images.<\/p><p>In contrast, in the panel for categories with more than 1000 images, we observe the genus and family taxa with about 40 labels each, making them good targets. Taxonomically, family is the least specific level, so we would expect more heterogeneity in images if we pick this level. Using genus seems then a good compromise between the taxonomic level and the number of images available.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e8b8b5b elementor-widget elementor-widget-heading\" data-id=\"e8b8b5b\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Final Dataset : Genus Taxa<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0a556ba elementor-widget elementor-widget-text-editor\" data-id=\"0a556ba\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The bar chart below shows the number of images for 41 selected genus (Fig. 5). We decided to limit ourselves to the 15 highlighted genera containing ~1,000 images to reduce computation time and avoid a biased representation. With this, we have a total of 18,412 images available for the classification\/modeling part of our project.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-ae7e7cd elementor-widget elementor-widget-image\" data-id=\"ae7e7cd\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"448\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-uv8wrQ30Blxf8D1YdRJPwA-1024x574.png\" class=\"attachment-large size-large wp-image-177924\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-uv8wrQ30Blxf8D1YdRJPwA-1024x574.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-uv8wrQ30Blxf8D1YdRJPwA-300x168.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-uv8wrQ30Blxf8D1YdRJPwA-768x431.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-uv8wrQ30Blxf8D1YdRJPwA.png 1213w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 5. Histogram showing the distribution of images in the 41 genus containing more than 1000 images and the 15 genus selected for the project analyses (in the red square). (Image by author)<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-4c28fb9 e-flex e-con-boxed e-con\" data-id=\"4c28fb9\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-4fa76dc elementor-widget elementor-widget-heading\" data-id=\"4fa76dc\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Classification using Neural Networks<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f59b903 elementor-widget elementor-widget-heading\" data-id=\"f59b903\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Architectures<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-aa51be5 elementor-widget elementor-widget-text-editor\" data-id=\"aa51be5\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>As mentioned before, this project aims to identify mushroom images, which is a typical image classification problem that is usually addressed by <strong>deep learning methods<\/strong>. To achieve our goal, we tested three models, detailed below.<\/p><ul><li><strong>LeNet<\/strong><\/li><\/ul><p>First, we tested our data on a <strong>CNN model<\/strong> with a LeNet architecture (Y. LeCun et al., 1998) that contains layers adapted for image processing. We tested this model with two convolutional layers.<\/p><p>With a value of 0.98% accuracy, the basic model does not seem suitable for our goal, so we did not pursue this model further.<\/p><ul><li><strong>VGG16<\/strong><\/li><\/ul><p>Next, we used the <strong>VGG16<\/strong> model (Simonyan and Zisserman, 2015), an image classification model with a Transfer Learning approach.<\/p><p>In the first step, our model had one classification layer with a Sequential model, two neural layers (<strong>Dense<\/strong>), and two regularization layers (<strong>Dropout<\/strong>).<\/p><p>The results obtained were encouraging, with an accuracy of about 70%, but the model suffered from overfitting. To counter this, we added a Dense layer and increased the Dropout value. The figure 6 shows the final architecture of the model.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-3d5eea8 elementor-widget elementor-widget-image\" data-id=\"3d5eea8\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"284\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ccuLJoXZLE7ySYxfB4hGjQ-1024x364.jpeg\" class=\"attachment-large size-large wp-image-177925\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ccuLJoXZLE7ySYxfB4hGjQ-1024x364.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ccuLJoXZLE7ySYxfB4hGjQ-300x107.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ccuLJoXZLE7ySYxfB4hGjQ-768x273.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-ccuLJoXZLE7ySYxfB4hGjQ.jpeg 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 6. Final VGG16 Architecture (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-36126d3 elementor-widget elementor-widget-text-editor\" data-id=\"36126d3\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Additionally, we used the <strong>EarlyStopping<\/strong> and <strong>ReduceLROnPlateau<\/strong> callbacks from the Keras package with patience values of 5 and 3 respectively. With this, we achieved an accuracy value of 75.5% without overfitting.<\/p><ul><li><strong>EfficientNetB1<\/strong><\/li><\/ul><p>The third image classification method we used in this project was also a <strong>Transfer Learning<\/strong> approach with the <strong>Tensorflow EfficientNet model<\/strong> (Tan and Le, 2019). We took inspiration from our previous model (VGG16) for the classification architecture we implemented, as we had positive results (fig. 7). We also used the callbacks <strong>EarlyStopping<\/strong> and <strong>ReduceLROnPlateau<\/strong> with the same values.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4ea4180 elementor-widget elementor-widget-image\" data-id=\"4ea4180\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"285\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-WR5X52mYnAYWvy5vS1Odag-1024x365.jpeg\" class=\"attachment-large size-large wp-image-177926\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-WR5X52mYnAYWvy5vS1Odag-1024x365.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-WR5X52mYnAYWvy5vS1Odag-300x107.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-WR5X52mYnAYWvy5vS1Odag-768x274.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-WR5X52mYnAYWvy5vS1Odag.jpeg 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 7. Fnal EfficientNetB1 architecture (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-ce4fd1e elementor-widget elementor-widget-text-editor\" data-id=\"ce4fd1e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The first results show an accuracy value of about 80%, better than with VGG16, but still showing overfitting, so we increased the value of the Dropout layers and finally got a score of 79% without overfitting.<\/p><p>Here is a display of the <strong>model\u2019s predictions<\/strong> on six images :<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-881d50a elementor-widget elementor-widget-image\" data-id=\"881d50a\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"796\" height=\"449\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-3_BydMeljxkUSkFguGyCEQ.png\" class=\"attachment-large size-large wp-image-177927\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-3_BydMeljxkUSkFguGyCEQ.png 796w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-3_BydMeljxkUSkFguGyCEQ-300x169.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-3_BydMeljxkUSkFguGyCEQ-768x433.png 768w\" sizes=\"(max-width: 796px) 100vw, 796px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 8. Example of prediction result, showing the predicted genus vs true genus. (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f469924 elementor-widget elementor-widget-heading\" data-id=\"f469924\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Error Analysis<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-01c8914 elementor-widget elementor-widget-text-editor\" data-id=\"01c8914\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<ul><li><strong>Confusion matrix<\/strong><\/li><\/ul><p>The confusion matrix (fig. 9) shows that classes 0, 3, and 8 tend to be classified as classes 7, 4, and 13, respectively :<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-eb53ad2 elementor-widget elementor-widget-image\" data-id=\"eb53ad2\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"616\" height=\"533\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-dyRBEhcoZ9ST5Po4rYDOOQ.png\" class=\"attachment-large size-large wp-image-177928\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-dyRBEhcoZ9ST5Po4rYDOOQ.png 616w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-dyRBEhcoZ9ST5Po4rYDOOQ-300x260.png 300w\" sizes=\"(max-width: 616px) 100vw, 616px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 9. Confusion matrix from EfficientNet prediction (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-df53ee5 elementor-widget elementor-widget-text-editor\" data-id=\"df53ee5\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Literature shows that these confused classes are very similar morphologically and belong to the same taxonomic family (table 3).<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d762807 elementor-widget elementor-widget-image\" data-id=\"d762807\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"308\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-JAkeplTY0KtrSQ3NTeepZQ-1024x394.png\" class=\"attachment-large size-large wp-image-177929\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-JAkeplTY0KtrSQ3NTeepZQ-1024x394.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-JAkeplTY0KtrSQ3NTeepZQ-300x116.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-JAkeplTY0KtrSQ3NTeepZQ-768x296.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-JAkeplTY0KtrSQ3NTeepZQ.png 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Table 3. Confused classes.<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0e37259 elementor-widget elementor-widget-text-editor\" data-id=\"0e37259\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The figure 10 show these morphological similarities between confused classes :<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e7257ff elementor-widget elementor-widget-image\" data-id=\"e7257ff\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"245\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-_v0_lXNVE-Ol8302C8mUxQ-1024x313.jpeg\" class=\"attachment-large size-large wp-image-177930\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-_v0_lXNVE-Ol8302C8mUxQ-1024x313.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-_v0_lXNVE-Ol8302C8mUxQ-300x92.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-_v0_lXNVE-Ol8302C8mUxQ-768x235.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-_v0_lXNVE-Ol8302C8mUxQ.jpeg 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 10. Morphological similarities between confused classes (source: Wikipwdia).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e9adff4 elementor-widget elementor-widget-text-editor\" data-id=\"e9adff4\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>One reason for the confusion of classes may be that the taxonomic identification of mushrooms is wrong. The images we use come from a web page where users classify images of mushrooms only by looking at the images and according to their experience and not according to <strong>an objective taxonomic identification protocol<\/strong>.<\/p><p>Considering the<strong> confusion matrix<\/strong> results, we wanted to evaluate our model without one of the confounded genera to see the influence of the data in our model, as show in the table 4.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5077689 elementor-widget elementor-widget-image\" data-id=\"5077689\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"316\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-OvM7bHWEPLUQuMtokEszsg-1024x405.png\" class=\"attachment-large size-large wp-image-177931\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-OvM7bHWEPLUQuMtokEszsg-1024x405.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-OvM7bHWEPLUQuMtokEszsg-300x119.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-OvM7bHWEPLUQuMtokEszsg-768x304.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-OvM7bHWEPLUQuMtokEszsg.png 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Table 4. In red the suppresed genus.<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-c856be8 elementor-widget elementor-widget-text-editor\" data-id=\"c856be8\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The accuracy values for the 12 genera (without the three eliminated) is 85%, 6% higher than the previous result (79%). This result suggests that we would probably increase our accuracy if we had had a more accurate data set of classifications to train the model.<\/p><ul><li><strong>Quality of the images<\/strong><\/li><\/ul><p>We have also noticed that the quality of the images is very diverse (Fig. 11). Indeed, some images contain elements that can prevent the model from identifying the mushroom correctly, such as images with a lot of noise around the mushroom to be classified (images 1 and 2). Images with no mushroom to classify (image 3). Images with more than one mushroom (image 4).<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b5e989a elementor-widget elementor-widget-image\" data-id=\"b5e989a\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"662\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-afiVjhVh7UNOs1ZPyF8qYQ-1024x847.png\" class=\"attachment-large size-large wp-image-177932\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-afiVjhVh7UNOs1ZPyF8qYQ-1024x847.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-afiVjhVh7UNOs1ZPyF8qYQ-300x248.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-afiVjhVh7UNOs1ZPyF8qYQ-768x635.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-afiVjhVh7UNOs1ZPyF8qYQ.png 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Fig. 11. Exemple of image quality (image by author).<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-224caec elementor-widget elementor-widget-heading\" data-id=\"224caec\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Wrapping Up<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-da0926b elementor-widget elementor-widget-text-editor\" data-id=\"da0926b\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>The summary table below (table 5) shows the accuracy values for the three methods used. We also calculated the \u00ab\u00a0<a href=\"https:\/\/scikit-learn.org\/stable\/modules\/generated\/sklearn.metrics.top_k_accuracy_score.html#sklearn.metrics.top_k_accuracy_score\">top-k accuracy classification score\u00a0\u00bb<\/a> of the package sklearn.metrics (Pedregosa et al., 2011.) with k = 2 and 3 to assess how often the actual category falls between the two or three most likely categories. According to the results, EfficientNet would be the most suitable approach for our mushroom image classification problem with 80% and 86% accuracy.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-df54a4c elementor-widget elementor-widget-image\" data-id=\"df54a4c\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t<figure class=\"wp-caption\">\n\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"800\" height=\"307\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-mpzE_M7_TTJPg79biOvTlg-1024x393.jpeg\" class=\"attachment-large size-large wp-image-177933\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-mpzE_M7_TTJPg79biOvTlg-1024x393.jpeg 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-mpzE_M7_TTJPg79biOvTlg-300x115.jpeg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-mpzE_M7_TTJPg79biOvTlg-768x295.jpeg 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/1-mpzE_M7_TTJPg79biOvTlg.jpeg 1400w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/>\t\t\t\t\t\t\t\t\t\t\t<figcaption class=\"widget-image-caption wp-caption-text\">Table 5. Summary of results.<\/figcaption>\n\t\t\t\t\t\t\t\t\t\t<\/figure>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9f405cf elementor-widget elementor-widget-text-editor\" data-id=\"9f405cf\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>Given the results, computer vision appears as an excellent alternative to the morphological classification of species, if the model could be trained using <strong>a large number of good quality images<\/strong> and that the species appearing in the image have a <strong>taxonomic classification made by a specialist<\/strong>.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9d31327 elementor-widget elementor-widget-heading\" data-id=\"9d31327\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h4 class=\"elementor-heading-title elementor-size-default\">Areas of improvement<\/h4>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bfef60e elementor-widget elementor-widget-text-editor\" data-id=\"bfef60e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p>These results can be improved in several ways. The first thing should be to train the model with an adequate <strong>dataset<\/strong>, i.e., better quality images and a more detailed taxonomic classification. Also, increasing the number of images in each label (e.g., from 1000 to 2000) could help to reduce morphological classification errors due to subjective human evaluation.<\/p><p>Another improvement will be increasing the number of genus. Currently, we trained the model using 15 of the 41 genus that had ~1000 images per label. Furthermore, training the same model to further identify images of mushrooms at the species level can also increase the accuracy.<\/p><p>Finally, combining the feature extraction part of our model with a classical classification model, such as <strong>Random Forest<\/strong> or <strong>SVM<\/strong>, could improve the classification.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-f34ac29 e-flex e-con-boxed e-con\" data-id=\"f34ac29\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-ce34f46 elementor-widget elementor-widget-heading\" data-id=\"ce34f46\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">References<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7741180 elementor-widget elementor-widget-text-editor\" data-id=\"7741180\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<ol class=\"\"><li id=\"a5dd\" class=\"nw nx iw ll b lm ln lq lr lu oi ly oj mc ok mg or oc od oe fy\" data-selectable-paragraph=\"\">Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner. 1998. \u00ab\u00a0Gradient-based learning applied to document recognition\u00a0\u00bb.Proceedings of the IEEE.<\/li><li id=\"7d7d\" class=\"nw nx iw ll b lm os lq ot lu ou ly ov mc ow mg or oc od oe fy\" data-selectable-paragraph=\"\">K. Simonyan, A. Zisserman. 2015. \u00ab\u00a0Very Deep Convolutional Networks for Large-Scale Image Recognition\u00a0\u00bb. Proceedings of the ICLR.<\/li><li id=\"6e70\" class=\"nw nx iw ll b lm os lq ot lu ou ly ov mc ow mg or oc od oe fy\" data-selectable-paragraph=\"\">Tan, Mingxing, and Quoc Le. 2019. \u00ab\u00a0EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks\u00a0\u00bb. Proceedings of Machine Learning Research 97.<\/li><li id=\"6d87\" class=\"nw nx iw ll b lm os lq ot lu ou ly ov mc ow mg or oc od oe fy\" data-selectable-paragraph=\"\">F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot and E. Duchesnay. 2011. \u00ab\u00a0Scikit-learn: Machine Learning in Python\u00a0\u00bb. Journal of Machine Learning Research.<\/li><\/ol>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Context One of the most widely used and oldest species identification techniques is \u00ab\u00a0morphological identification\u00a0\u00bb, which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it. An alternative to this identification technique [&hellip;]<\/p>\n","protected":false},"author":77,"featured_media":177888,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2497],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Mushroom recognition: How computer vision can aid in the identification of species<\/title>\n<meta name=\"description\" content=\"One of the most widely used and oldest species identification techniques is \u201cmorphological identification,\u201d which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Mushroom recognition : How computer vision can aid in the identification of species\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2022-06-01T08:54:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T14:04:06+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/mushroom-identification-computer-vision.png\" \/>\n\t<meta property=\"og:image:width\" content=\"900\" \/>\n\t<meta property=\"og:image:height\" content=\"450\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Raphael Kassel\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"Raphael Kassel\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"13 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species\",\"url\":\"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species\",\"name\":\"Mushroom recognition: How computer vision can aid in the identification of species\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2022-06-01T08:54:00+00:00\",\"dateModified\":\"2023-11-09T14:04:06+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\"},\"description\":\"One of the most widely used and oldest species identification techniques is \u201cmorphological identification,\u201d which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it.\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\",\"name\":\"Raphael Kassel\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"caption\":\"Raphael Kassel\"},\"description\":\"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \\\"Chef de projet IA\\\" et \\\"Data Product Manager\\\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/raphaelkassel\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/raphaelkassel\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Mushroom recognition: How computer vision can aid in the identification of species","description":"One of the most widely used and oldest species identification techniques is \u201cmorphological identification,\u201d which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it.","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:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","og_locale":"fr_FR","og_type":"article","og_title":"Mushroom recognition : How computer vision can aid in the identification of species","og_url":"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2022-06-01T08:54:00+00:00","article_modified_time":"2023-11-09T14:04:06+00:00","og_image":[{"width":900,"height":450,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2022\/06\/mushroom-identification-computer-vision.png","type":"image\/png"}],"author":"Raphael Kassel","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"Raphael Kassel","Dur\u00e9e de lecture estim\u00e9e":"13 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","url":"https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species","name":"Mushroom recognition: How computer vision can aid in the identification of species","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2022-06-01T08:54:00+00:00","dateModified":"2023-11-09T14:04:06+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1"},"description":"One of the most widely used and oldest species identification techniques is \u201cmorphological identification,\u201d which identifies individuals by their anatomical characteristics. However, this technique has the disadvantage of not being always accurate since it depends on the observation and the identification protocol of the person who performs it.","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/mushroom-recognition-how-computer-vision-can-aid-in-the-identification-of-species"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1","name":"Raphael Kassel","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","caption":"Raphael Kassel"},"description":"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \"Chef de projet IA\" et \"Data Product Manager\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.","sameAs":["https:\/\/www.linkedin.com\/in\/raphaelkassel\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/raphaelkassel"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/164614"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/77"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=164614"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/164614\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/177888"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=164614"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=164614"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=164614"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}