{"id":22941096,"url":"https://github.com/narius2030/imcp-support-blinders","last_synced_at":"2026-06-30T18:31:14.383Z","repository":{"id":267902807,"uuid":"897253549","full_name":"Narius2030/IMCP-Support-Blinders","owner":"Narius2030","description":"This project focuses on image captioning by creating two primary models: DarkNetLM and DarkNetVG2. Both models leverage the CSP DarkNet53 architecture as the backbone of YOLOv8 for feature extraction from images. Combining with Transformers or LSTM to generating captions.","archived":false,"fork":false,"pushed_at":"2025-06-09T15:50:52.000Z","size":172625,"stargazers_count":1,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-10-30T17:50:28.081Z","etag":null,"topics":["computer-vision","data-lake","image-captioning","large-language-model","mobile-app"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","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/Narius2030.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,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2024-12-02T10:06:47.000Z","updated_at":"2025-06-26T08:52:36.000Z","dependencies_parsed_at":null,"dependency_job_id":"b2facd58-bd12-4392-b943-66d9c3b2b5c4","html_url":"https://github.com/Narius2030/IMCP-Support-Blinders","commit_stats":null,"previous_names":["narius2030/imcp-support-blinders"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Narius2030/IMCP-Support-Blinders","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Narius2030%2FIMCP-Support-Blinders","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Narius2030%2FIMCP-Support-Blinders/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Narius2030%2FIMCP-Support-Blinders/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Narius2030%2FIMCP-Support-Blinders/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Narius2030","download_url":"https://codeload.github.com/Narius2030/IMCP-Support-Blinders/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Narius2030%2FIMCP-Support-Blinders/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34979577,"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":"2026-06-30T02:00:05.919Z","response_time":92,"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":["computer-vision","data-lake","image-captioning","large-language-model","mobile-app"],"created_at":"2024-12-14T13:36:40.869Z","updated_at":"2026-06-30T18:31:14.377Z","avatar_url":"https://github.com/Narius2030.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🚦 Image Caption Support Blinder\n\n## 📑 Table of Contents\n\n- [🚦 Image Caption Support Blinder](#-image-caption-support-blinder)\n  - [📑 Table of Contents](#-table-of-contents)\n  - [📝 Overview](#-overview)\n  - [📁 Folder Structure](#-folder-structure)\n  - [🔗 Pipeline](#-pipeline)\n  - [🤖 Models](#-models)\n  - [🚀 Usage](#-usage)\n    - [1. Environment Setup](#1-environment-setup)\n    - [2. Run the Pipeline Steps](#2-run-the-pipeline-steps)\n  - [📊 Results](#-results)\n  - [🤝 Contributing](#-contributing)\n  - [🪪 License](#-license)\n  - [🙏 Acknowledgments](#-acknowledgments)\n  - [💁‍♂️ Author](#️-author)\n\n## 📝 Overview\n\nThis project provides a full pipeline for building a Vietnamese traffic image captioning dataset and system, supporting both computer vision research and accessibility for the visually impaired. The pipeline includes data crawling, cleaning, caption generation, and image augmentation.\n\n## 📁 Folder Structure\n\n```\nIMCP-Support-Blinders/\n├── README.md\n├── Data/\n│   ├── dashboard.dio\n│   ├── image.png\n│   ├── README_en.md\n│   ├── README_vn.md\n│   ├── README.md\n│   ├── 1.crawl_data/\n│   │   ├── env_crawl_data.yaml\n│   │   ├── output/\n│   │   │   ├── metadata.json\n│   │   │   ├── traffic_images_dataset_v1.csv\n│   │   │   ├── traffic_images_dataset_v2.csv\n│   │   │   └── traffic_images_dataset_v3.csv\n│   │   └── python/\n│   │       └── traffic_raw.py\n│   ├── 2.data_preprocessing/\n│   │   ├── jupyter/\n│   │   │   └── data_preprocessing.ipynb\n│   │   └── output/\n│   ├── 3.labels_short_captions/\n│   │   ├── output/\n│   │   │   └── csv_with_captions/\n│   │   └── python/\n│   └── 4.Image_data_augument/\n│       ├── output/\n│       └── python/\n├── Model/\n│   ├── DarkNetLM/\n│   └── DarkNetVG2/\n```\n\n- **Data/**: Main pipeline and dataset scripts, including crawling, preprocessing, captioning, and augmentation.\u003cbr\u003e\n  👉 For more details, see the [Germini-Captioning-Dataset-2025 repo](https://github.com/TrieuPhi/Germini-Captioning-Dataset-2025)\n- **Model/**: Model code and experiments (details to be updated).\n\n## 🔗 Pipeline\n\n1. **Crawl Data**  \n   Collect traffic images from Google Images via SerpApi, save metadata and images locally.\n\n2. **Data Preprocessing**  \n   Clean data: remove broken URLs, handle nulls, standardize fields.\n\n3. **Labels Short Captions**  \n   Use Gemini 2.0 Flash API to generate concise (10-15 words) captions for each image.\n\n4. **Image Augmentation**  \n   Augment data using modern image transformation techniques with [Albumentations](https://albumentations.ai/).\n\n## 🤖 Models\n\n**To be updated.**\n\n## 🚀 Usage\n\n### 1. Environment Setup\n\n```bash\ncd Data/1.crawl_data/\nconda env create -f env_crawl_data.yaml\nconda activate crawl_data\npip install -r ../3.labels_short_captions/python/requirements.txt\npip install -r ../4.Image_data_augument/python/requirements.txt\n```\n\n### 2. Run the Pipeline Steps\n\n**Step 1: Crawl data**\n\n```bash\npython 1.crawl_data/python/traffic_raw.py\n```\n\n**Step 2: Data cleaning**\n\n- Use the notebook or script in `2.data_preprocessing/jupyter/`\n\n**Step 3: Caption generation**\n\n```bash\npython 3.labels_short_captions/python/label_short_captions.py\n```\n\n**Step 4: Augmentation**\n\n```bash\npython 4.Image_data_augument/python/data_augument.py\n```\n\n## 📊 Results\n\n- Augmented images and captions are stored in `Data/4.Image_data_augument/output/` and `augmented/`.\n- Cleaned CSVs and intermediate results are in the corresponding `output/` folders.\n\n## 🤝 Contributing\n\nContributions are welcome! Please submit a pull request or open an issue for any suggestions or improvements.\n\n## 🪪 License\n\nThis project is licensed under the MIT License. See the LICENSE file for more details.\n\n## 🙏 Acknowledgments\n\n- BERT, GPT-2, and Albumentations authors.\n\n## 💁‍♂️ Author\n\n- [Narius2030](https://github.com/Narius2030) 🦉\n- [HTN-DT-Beo](https://github.com/HTN-DT-Beo) 🐻\n- [TrieuPhi](https://github.com/TrieuPhi) 🚀\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnarius2030%2Fimcp-support-blinders","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnarius2030%2Fimcp-support-blinders","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnarius2030%2Fimcp-support-blinders/lists"}