{"id":18160836,"url":"https://github.com/blankeos/cgjj-detector","last_synced_at":"2026-06-25T08:31:42.662Z","repository":{"id":108680398,"uuid":"571857107","full_name":"Blankeos/cgjj-detector","owner":"Blankeos","description":"🕵️‍♀️ The CGJJ or the Carlo Glecy Jessa Jonah Detector is an object detection model trained on YoloV5 to detect the faces of Carlo, Glecy, Jessa, and Jonah.","archived":false,"fork":false,"pushed_at":"2022-12-05T05:44:18.000Z","size":43679,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-04-07T03:54:52.547Z","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/Blankeos.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":"2022-11-29T03:00:52.000Z","updated_at":"2022-11-30T01:57:10.000Z","dependencies_parsed_at":"2023-05-17T08:40:47.562Z","dependency_job_id":null,"html_url":"https://github.com/Blankeos/cgjj-detector","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Blankeos/cgjj-detector","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Blankeos%2Fcgjj-detector","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Blankeos%2Fcgjj-detector/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Blankeos%2Fcgjj-detector/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Blankeos%2Fcgjj-detector/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Blankeos","download_url":"https://codeload.github.com/Blankeos/cgjj-detector/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Blankeos%2Fcgjj-detector/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34767542,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-25T02:00:05.521Z","response_time":101,"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":"2024-11-02T08:09:30.483Z","updated_at":"2026-06-25T08:31:42.640Z","avatar_url":"https://github.com/Blankeos.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🕵️‍♀️ CGJJ Detector\n\n![cgjj preview](/docs/cgjj-preview.gif)\n\n### 🤔 About\n\nThe **CGJJ** or the **Carlo Glecy Jessa Jonah Detector** is an object detection model trained on YoloV5 to detect the faces of Carlo, Glecy, Jessa, and Jonah.\n\nWe annotated a custom 30FPS video dataset on [makesense.ai](https://makesense.ai/) for an object detection task of 4 classes: `carlo`, `glecy`, `jessa`, `jj`. We have 665 images for training and 100 images for testing.\n\nThis repository has two purposes:\n\n1. [Use the model for inference](#🚀-get-started)\n2. [Recreate the model from scratch](#😎-how-to-recreate-this-app-from-scratch)\n\nSubmitted to **Mr. John Christopher Mateo** for **CCS 250 Computer Vision** as a Final Project for the semester.\n\n### 🚀 How to Use:\n\n1. Clone this repository\n\n```sh\n$ git clone https://www.github.com/blankeos/cgjj-detector\n$ cd cgjj-detector\n```\n\n2. Install the requirements\n\n```sh\n$ python -m venv venv # Optional: Create a virtual environment\n$ pip install -r requirements.txt\n```\n\n3. Run scripts to use model for inference (You have 3 choices):\n\n   - [x] Create video from `./videoData/test.mp4` (Smoothest but Long Process Time)\n\n   ```sh\n   python create_vid.py\n   # Saves video on 'output_video.mp4'\n   ```\n\n   - [x] Use model in real-time on **camera** (Choppy depending on GPU)\n\n   ```sh\n   python realtime_cam.py\n   ```\n\n   - [x] Use model in real-time on `./videoData/test.mp4` (Choppy depending on GPU)\n\n   ```sh\n   python realtime_vid.py\n   ```\n\n### 📝 Important Links\n\n- [Final Annotated Dataset](https://drive.google.com/drive/folders/18rj0MZ2vT_22wnRXBmll7RkDSqnGJXXE?usp=share_link)\n- [YoloV5 Training Notebook](https://colab.research.google.com/drive/1qct0OJryhOWcNPBE15wgIgECriVE5X75?usp=sharing)\n- [Final Model Used](/model/cgjj_best.pt)\n\n### 📁 Dataset Directory Structure\n\n```\n| - dataset\n    | -- images\n        | -- train\n        | -- val\n    | -- labels\n        | -- train\n        | -- val\n    labels.txt\n```\n\n### 🧠 Final Model Performance\n\n```\nModel summary: 157 layers, 7020913 parameters, 0 gradients, 15.8 GFLOPs\n                 Class     Images  Instances          P          R      mAP50   mAP50-95: 100% 4/4 [00:01\u003c00:00,  2.30it/s]\n                   all        100        400      0.996      0.997      0.995      0.775\n                 carlo        100        100          1       0.99      0.995      0.789\n                 glecy        100        100      0.998          1      0.995      0.689\n                 jessa        100        100      0.999          1      0.995      0.815\n                    jj        100        100      0.988          1      0.995      0.805\n```\n\n---\n\n### 😎 How to recreate this project from scratch\n\nMake sure you already cloned this repository and installed the requirements.\n\n1. Run preprocessing script\n\n```sh\npython preprocessing.py\n# This will use `./videoData` and create images following the YoloV5 directory structure\n# Then, outputs the preprocessed data on `/cgjj-dataset`\n```\n\n2. Annotate the dataset on [makesense.ai](https://makesense.ai/) for an object detection task of 4 classes: `carlo`, `glecy`, `jessa`, `jj`. _(It took us 6 hours to do this)_ 😅\n3. Upload the dataset on your Google Drive (To be trained on Colab).\n4. Train the dataset on our **YoloV5 Training Notebook**.\n5. Download the model from `runs/expt/best_weights.pt` in Colab\n6. Put the `best_weights.pt` inside `/model` dir of this repository. Rename as `cgjj_best.pt`.\n7. Go back to Step 3 in [Get Started](#🚀-get-started).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fblankeos%2Fcgjj-detector","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fblankeos%2Fcgjj-detector","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fblankeos%2Fcgjj-detector/lists"}