{"id":18585376,"url":"https://github.com/gjy3035/pcc-net","last_synced_at":"2025-04-10T13:30:58.399Z","repository":{"id":134756423,"uuid":"162512887","full_name":"gjy3035/PCC-Net","owner":"gjy3035","description":"PCC Net: Perspective Crowd Counting via Spatial Convolutional Network","archived":false,"fork":false,"pushed_at":"2020-03-12T11:12:23.000Z","size":692,"stargazers_count":76,"open_issues_count":18,"forks_count":18,"subscribers_count":6,"default_branch":"ori_pt1_py3","last_synced_at":"2025-03-24T21:39:12.249Z","etag":null,"topics":["computer-vision","crowd-analysis","crowd-counting","multi-task-learning"],"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/gjy3035.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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2018-12-20T02:00:13.000Z","updated_at":"2025-01-21T06:34:03.000Z","dependencies_parsed_at":null,"dependency_job_id":"64f17c5d-6156-4d42-a477-ed9a7617ff13","html_url":"https://github.com/gjy3035/PCC-Net","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gjy3035%2FPCC-Net","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gjy3035%2FPCC-Net/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gjy3035%2FPCC-Net/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gjy3035%2FPCC-Net/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/gjy3035","download_url":"https://codeload.github.com/gjy3035/PCC-Net/tar.gz/refs/heads/ori_pt1_py3","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248225707,"owners_count":21068078,"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":["computer-vision","crowd-analysis","crowd-counting","multi-task-learning"],"created_at":"2024-11-07T00:33:38.671Z","updated_at":"2025-04-10T13:30:58.373Z","avatar_url":"https://github.com/gjy3035.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# PCC Net: Perspective Crowd Counting via Spatial Convolutional Network\nThis is an official implementation of the paper \"PCC net\" (PCC Net: Perspective Crowd Counting via Spatial Convolutional Network).\n\n![PCC Net.](./imgs/img0.png \"pcc\")\n\nIn the paper, the experiments are conducted on the three populuar datasets: Shanghai Tech, UCF_CC_50 and WorldExpo'10. To be specific, Shanghai Tech Part B contains crowd images with the same resolution. For easier data prepareation, we only release the pre-trained model on ShanghaiTech Part B dataset in this repo.\n\n## Bracnhes\n\n1. [ori_pt0.2_py2](https://github.com/gjy3035/PCC-Net/tree/ori_pt0.2_py2): the original version.\n2. [ori_pt1_py3](https://github.com/gjy3035/PCC-Net): the current version.\n3. [vgg_pt1_py3](https://github.com/gjy3035/PCC-Net/tree/vgg_pt1_py3): vgg-backbone PCC Net (higher performance).\n\n##  Requirements\n- Python 3.x\n- Pytorch 1.x\n- TensorboardX (pip)\n- torchvision  (pip)\n- easydict (pip)\n- pandas  (pip)\n\n\n## Data preparation\n1. Download the original ShanghaiTech Dataset [Link: [Dropbox ](https://www.dropbox.com/s/fipgjqxl7uj8hd5/ShanghaiTech.zip?dl=0)/ [BaiduNetdisk](https://pan.baidu.com/s/1nuAYslz)]\n2. Resize the images and the locations of key points. \n3. Generate the density maps by using the [code](https://github.com/aachenhang/crowdcount-mcnn/tree/master/data_preparation).\n4. Generate the segmentation maps.\n\nWe also provide the processed Part B dataset for training. [[Link](https://mailnwpueducn-my.sharepoint.com/:u:/g/personal/gjy3035_mail_nwpu_edu_cn/EcMLqr9XuH1ChAgkqpxL_6kBK9EyCmIuXMxTb09FrjMYow?e=LJnOcC)]\n\n## Training model\n1. Run the train_lr.py: ```python train_lr.py```.\n2. See the training outputs: ```Tensorboard --logdir=exp --port=6006```.\n\nIn the experiments, training  and tesing 800 epoches take 21 hours on GTX 1080Ti. \n\n## Expermental results\n\n### Quantitative results\n\nWe show the Tensorboard visualization results as below:\n![Detialed infomation during the traning phase.](./imgs/img1.jpg \"pcc_q\")\nThe mae and mse are the results on test set. Others are triaining loss. \n\n### Visualization results\nVisualization results on the test set as below:\n![Visualization results on the test set.](./imgs/img2.jpg \"pcc_v\")\nColumn 1: input image; Column 2: density map GT; Column 3: density map prediction; Column 4: segmentation map GT; Column 5: segmentation map prediction.\n\n## Citation\nIf you use the code, please cite the following paper:\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgjy3035%2Fpcc-net","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fgjy3035%2Fpcc-net","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgjy3035%2Fpcc-net/lists"}