{"id":26486587,"url":"https://github.com/riciokzz/computer-vision","last_synced_at":"2026-05-20T04:46:40.045Z","repository":{"id":283344426,"uuid":"951468432","full_name":"Riciokzz/Computer-Vision","owner":"Riciokzz","description":"Computer Vision project","archived":false,"fork":false,"pushed_at":"2025-03-19T18:16:26.000Z","size":830,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-19T19:27:18.817Z","etag":null,"topics":["cuda","data-cleaning","data-engineering","data-science","exploratory-data-analysis","machine-learning","neural-network"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Riciokzz.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2025-03-19T18:15:53.000Z","updated_at":"2025-03-19T18:19:05.000Z","dependencies_parsed_at":"2025-03-20T02:46:28.547Z","dependency_job_id":null,"html_url":"https://github.com/Riciokzz/Computer-Vision","commit_stats":null,"previous_names":["riciokzz/computer-vision"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Riciokzz%2FComputer-Vision","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Riciokzz%2FComputer-Vision/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Riciokzz%2FComputer-Vision/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Riciokzz%2FComputer-Vision/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Riciokzz","download_url":"https://codeload.github.com/Riciokzz/Computer-Vision/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244560839,"owners_count":20472298,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cuda","data-cleaning","data-engineering","data-science","exploratory-data-analysis","machine-learning","neural-network"],"created_at":"2025-03-20T06:19:53.794Z","updated_at":"2026-05-20T04:46:40.039Z","avatar_url":"https://github.com/Riciokzz.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Computer Vision\n\n## Introduction\n\nIn US alone, around 7500 yearly cases of mushrooms poisoning are \nreported [SOURCE](https://www.tandfonline.com/doi/full/10.1080/00275514.2018.1479561). According to the source, \n\"misidentification of edible mushroom species appears to be the most common cause \nand may be preventable through education\". To avoid expenses for hospitalization \nand in some cases pointless deaths, I have been hired by US National Health \nService to create a machine learning model, that can recognize mushroom types. \nThey want to install this on hand-held devices and to help people make the \nright choice when mushroom picking.\n\n\nObjectives of this project\n\u003cul\u003e\n    \u003cli\u003eBuild Data Pipeline\u003c/li\u003e\n    \u003cli\u003ePreprocess Data\u003c/li\u003e\n    \u003cli\u003eBuild Deep Neural Network\u003c/li\u003e\n    \u003cli\u003eEvaluate Performance\u003c/li\u003e\n\u003c/ul\u003e\n\nThe aim of the Project\nBuild DNN model which help to identify Edible and Lethal poisonous mushrooms using transfer learning. This project will include data preprocessing, data analysis, model building steps.\n\nThe main dataset can be downloaded from [Kaggle](https://www.kaggle.com/maysee/mushrooms-classification-common-genuss-images)\n\n\n## Requirements for the project\n\nTo install all necessary libraries use: `pip install -r requirements.txt`\n\nMushrooms folder with all subclasses should be put in the `data` folder.\n\n![img.png](images/structure_data.png)\n\n## Summary\n\nWe found some images which are not usable for NN model and analysis, \nthey totally white or corrupted and part of image not loaded. Those images move to \nfolder bad and will not be used.\n\nMushroom dataset consists of 9 different classes, dataset is imbalance. \n\nMost of the images takes from 50KB to 200KB. \n\nWe worked with of 6711 images, as number is quite low we use image augmentation - \nrotation, flipping, brightness adjustment to create additional images.\n\nMost of the images are 800x580 size.\n\nWe created a Neural Network model using backbone of ResNet18 model.\n\u003cul\u003e\n    \u003cli\u003eUse CUDA for faster learning\u003c/li\u003e\n    \u003cli\u003eFine Tune model\u003c/li\u003e\n    \u003cli\u003eSet different learning rate for head of model and backbone\u003c/li\u003e\n    \u003cli\u003eFor head use smaller epoch count, increase it while unfreezing the \nbackbone of the model\u003c/li\u003e\n    \u003cli\u003eSaved the best model with the lowest validation loss -\u003e best_model_checkpoint_*.ckpt\u003c/li\u003e\n    \u003cli\u003eModel evaluated\u003c/li\u003e\n    \u003cli\u003eAccuracy score with unseen data: 87%\u003c/li\u003e\n\u003c/ul\u003e\n\n\n## License\nThis project is licensed under the MIT License - see the LICENSE file for details.\n\n## Contact Information\n[Email](ricardas.poskrebysev@gmail.com)\n[LinkedIn](https://www.linkedin.com/in/ri%C4%8Dardas-poskreby%C5%A1evas-665207206/)\n[GitHub](https://github.com/Riciokzz)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Friciokzz%2Fcomputer-vision","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Friciokzz%2Fcomputer-vision","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Friciokzz%2Fcomputer-vision/lists"}