{"id":28479711,"url":"https://github.com/dizcza/ujipen","last_synced_at":"2025-10-11T15:42:02.114Z","repository":{"id":84361038,"uuid":"171858288","full_name":"dizcza/ujipen","owner":"dizcza","description":"UJIPEN2 classification with Gated Recurrent Unit Neural Network.","archived":false,"fork":false,"pushed_at":"2024-05-03T20:11:06.000Z","size":52,"stargazers_count":1,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-09-27T05:55:11.549Z","etag":null,"topics":["dtw","gru","handwriting-recognition"],"latest_commit_sha":null,"homepage":"","language":"Python","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/dizcza.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":"2019-02-21T11:19:50.000Z","updated_at":"2020-09-11T18:51:36.000Z","dependencies_parsed_at":"2023-03-12T22:29:08.584Z","dependency_job_id":null,"html_url":"https://github.com/dizcza/ujipen","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/dizcza/ujipen","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dizcza%2Fujipen","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dizcza%2Fujipen/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dizcza%2Fujipen/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dizcza%2Fujipen/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/dizcza","download_url":"https://codeload.github.com/dizcza/ujipen/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dizcza%2Fujipen/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279007601,"owners_count":26084334,"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","status":"online","status_checked_at":"2025-10-11T02:00:06.511Z","response_time":55,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["dtw","gru","handwriting-recognition"],"created_at":"2025-06-07T18:10:08.358Z","updated_at":"2025-10-11T15:42:02.094Z","avatar_url":"https://github.com/dizcza.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# UjipenChars2 handwritten letters classifier with Gated Recurrent Unit (GRU)\n\nThis repository supplements [stm32f429-chars](https://github.com/dizcza/stm32f429-chars) repository to train a recurrent neural network that will be used later on in a microcontroller.\n\nA small list of manually picked examples from train data which confuse classifiers is put in [`dropped.txt`](ujipenchars2/dropped.txt). All test samples from [UjipenChars2](https://archive.ics.uci.edu/ml/datasets/UJI+Pen+Characters+(Version+2)) dataset are used during the model validation.\n\nThe main file is [`gru.py`](gru.py), where the training procedure of GRU is defined alongside with the test (validation) score.\n\nInitially started with DTW as a baseline algorithm to find the closest pattern from the train data, given an input sample. DTW-related implementation is moved to [dtw](https://github.com/dizcza/ujipen/tree/dtw) branch.\n\nTo give you the rough approximation of performance of both classifiers,\n\n|                     |  GRU   |  DTW   |\n|---------------------|--------|--------|\n| Validation accuracy | 98.3 % | 81.9 % |\n\nBut the main difference between those two is their inference time: GRU is much faster than DTW due to parallel computation.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdizcza%2Fujipen","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdizcza%2Fujipen","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdizcza%2Fujipen/lists"}