{"id":32986051,"url":"https://github.com/WeidiXie/cell_counting_v2","last_synced_at":"2025-11-18T02:02:08.066Z","repository":{"id":49325148,"uuid":"86938424","full_name":"WeidiXie/cell_counting_v2","owner":"WeidiXie","description":null,"archived":false,"fork":false,"pushed_at":"2017-05-30T16:10:46.000Z","size":11,"stargazers_count":119,"open_issues_count":1,"forks_count":45,"subscribers_count":11,"default_branch":"master","last_synced_at":"2024-07-24T04:34:52.210Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/WeidiXie.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2017-04-01T19:46:24.000Z","updated_at":"2024-06-06T23:32:58.000Z","dependencies_parsed_at":"2022-09-23T11:00:47.691Z","dependency_job_id":null,"html_url":"https://github.com/WeidiXie/cell_counting_v2","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/WeidiXie/cell_counting_v2","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/WeidiXie%2Fcell_counting_v2","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/WeidiXie%2Fcell_counting_v2/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/WeidiXie%2Fcell_counting_v2/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/WeidiXie%2Fcell_counting_v2/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/WeidiXie","download_url":"https://codeload.github.com/WeidiXie/cell_counting_v2/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/WeidiXie%2Fcell_counting_v2/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":284988465,"owners_count":27095952,"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":"2025-11-18T02:00:05.759Z","response_time":61,"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":[],"created_at":"2025-11-13T08:00:36.460Z","updated_at":"2025-11-18T02:02:08.061Z","avatar_url":"https://github.com/WeidiXie.png","language":"Python","readme":"# cell_counting_v2\n\nThe repository includes the code for training cell counting applications. (Keras + Tensorflow)\n\nDataset can be downloaded here :\nhttp://www.robots.ox.ac.uk/~vgg/research/counting/index_org.html\n\nRelated Papers:\n\n[1] Microscopy Cell Counting with Fully Convolutional Regression Networks.\n\nhttp://www.robots.ox.ac.uk/~vgg/publications/2015/Xie15/weidi15.pdf\n\n[2] U-Net: Convolutional Networks for Biomedical Image Segmentation.\n\nhttps://arxiv.org/abs/1505.04597\n\nTo make the training easier, I added Batch Normalization to all architectures (FCRN-A and U-Net simple version).\n\nThough still contains tiny difference with the original Matconvnet implementation, \nfor instance, upsampling in Keras is implemented by repeating elements, instead of bilinear upsampling. \nSo, to mimic the bilinear upsampling, I did upsampling + convolution. \nAlso, more data augmentation needs to be added. \nNevertheless. I'm able to get similar results as reported in the paper.\n\nIn all architectures, they follow the fully convolutional idea, \neach architecture consists of a down-sampling path, followed by an up-sampling path. \nDuring the first several layers, the structure resembles the cannonical classification CNN, as convolution,\nReLU, and max pooling are repeatedly applied to the input image and feature maps. \nIn the second half of the architecture, spatial resolution is recovered by performing up-sampling, convolution, eventually mapping the intermediate feature representation back to the original resolution. \n\nIn the U-net version, low-level feature representations are fused during upsampling, aiming to compensate the information loss due to max pooling. Here, I only gave a very simple example here (64 kernels for all layers), not tuned for any dataset.\n\nAs people know, Deep Learning is developing extremely fast today, both papers were published two years ago,\nwhich is quite \"old\". If people are interested in cell counting, feel free to edit on this.\n\n\n\n\n","funding_links":[],"categories":["Pedestrain/Crowd"],"sub_categories":["3D SemanticSeg"],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FWeidiXie%2Fcell_counting_v2","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FWeidiXie%2Fcell_counting_v2","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FWeidiXie%2Fcell_counting_v2/lists"}