{"id":26610304,"url":"https://github.com/divith123/deepsecure-ai","last_synced_at":"2026-03-08T14:40:09.845Z","repository":{"id":264494039,"uuid":"890724896","full_name":"Divith123/DeepSecure-AI","owner":"Divith123","description":"DeepSecure-AI is an open-source deepfake detection tool leveraging advanced AI models like EfficientNetV2 and MTCNN for high-accuracy analysis of images, videos, and audios. With real-time frame-by-frame video processing and a user-friendly Gradio interface, it empowers researchers and developers to combat media falsification effectively.","archived":false,"fork":false,"pushed_at":"2025-07-12T05:36:24.000Z","size":1488,"stargazers_count":7,"open_issues_count":0,"forks_count":4,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-07-12T07:18:46.963Z","etag":null,"topics":["ai-models","deepfake-detection","machine-learning","python"],"latest_commit_sha":null,"homepage":"https://huggingface.co/spaces/ninjaonsteroids/DeepSecure-AI","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/Divith123.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":"2024-11-19T04:18:56.000Z","updated_at":"2025-07-12T05:36:28.000Z","dependencies_parsed_at":"2025-03-24T01:39:10.128Z","dependency_job_id":null,"html_url":"https://github.com/Divith123/DeepSecure-AI","commit_stats":null,"previous_names":["divith123/deepsecure-ai"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Divith123/DeepSecure-AI","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Divith123%2FDeepSecure-AI","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Divith123%2FDeepSecure-AI/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Divith123%2FDeepSecure-AI/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Divith123%2FDeepSecure-AI/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Divith123","download_url":"https://codeload.github.com/Divith123/DeepSecure-AI/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Divith123%2FDeepSecure-AI/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":267141156,"owners_count":24041982,"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":"2025-07-26T02:00:08.937Z","response_time":62,"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":["ai-models","deepfake-detection","machine-learning","python"],"created_at":"2025-03-24T01:39:03.235Z","updated_at":"2026-03-08T14:40:04.823Z","avatar_url":"https://github.com/Divith123.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# DeepSecure-AI\n\nDeepSecure-AI is a powerful open-source tool designed to detect fake images, videos, and audios. Utilizing state-of-the-art deep learning techniques like EfficientNetV2 and MTCNN, DeepSecure-AI offers frame-by-frame video analysis, enabling high-accuracy deepfake detection. It's developed with a focus on ease of use, making it accessible for researchers, developers, and security analysts...\n\n---\n\n## Features\n\n- Multimedia Detection: Detect deepfakes in images, videos, and audio files using a unified platform.\n- High Accuracy: Leverages EfficientNetV2 for enhanced prediction performance and accurate results.\n- Real-Time Video Analysis: Frame-by-frame analysis of videos with automatic face detection.\n- User-Friendly Interface: Easy-to-use interface built with Gradio for uploading and processing media files.\n- Open Source: Completely open source under the MIT license, making it available for developers to extend and improve.\n\n---\n\n## Demo-Data\n\nYou can test the deepfake detection capabilities of DeepSecure-AI by uploading your video files. The tool will analyze each frame of the video, detect faces, and determine the likelihood of the video being real or fake.\n\nExamples:  \n1. [Video1-fake-1-ff.mp4](#)\n2. [Video6-real-1-ff.mp4](#)\n\n---\n\n## How It Works\n\nDeepSecure-AI uses the following architecture:\n\n1. Face Detection:  \n   The [MTCNN](https://arxiv.org/abs/1604.02878) model detects faces in each frame of the video. If no face is detected, it will use the previous frame's face to ensure accuracy.\n\n2. Fake vs. Real Classification:  \n   Once the face is detected, it's resized and fed into the [EfficientNetV2](https://arxiv.org/abs/2104.00298) deep learning model, which determines the likelihood of the frame being real or fake.\n\n3. Fake Confidence:  \n   A final prediction is generated as a percentage score, indicating the confidence that the media is fake.\n\n4. Results:  \n   DeepSecure-AI provides an output video, highlighting the detected faces and a summary of whether the input is classified as real or fake.\n\n---\n\n## Project Setup\n\n### Prerequisites\n\nEnsure you have the following installed:\n\n- Python 3.10\n- Gradio (pip install gradio)\n- TensorFlow (pip install tensorflow)\n- OpenCV (pip install opencv-python)\n- PyTorch (pip install torch torchvision torchaudio)\n- facenet-pytorch (pip install facenet-pytorch)\n- MoviePy (pip install moviepy)\n\n### Installation\n\n1. Clone the repository:\n        git clone https://github.com/Divith123/DeepSecure-AI.git\n    cd DeepSecure-AI\n    \n\n2. Install required dependencies:\n        pip install -r requirements.txt\n    \n\n3. Download the pre-trained model weights for EfficientNetV2 and place them in the project folder.\n\n### Running the Application\n\n1. Launch the Gradio interface:\n        python app.py\n    \n\n2. The web interface will be available locally. You can upload a video, and DeepSecure-AI will analyze and display results.\n\n---\n\n## Example Usage\n\nUpload a video or image to DeepSecure-AI to detect fake media. Here are some sample predictions:\n\n- Video Analysis: The tool will detect faces from each frame and classify whether the video is fake or real.\n- Result Output: A GIF or MP4 file with the sequence of detected faces and classification result will be provided.\n\n---\n\n## Technologies Used\n\n- TensorFlow: For building and training deep learning models.\n- EfficientNetV2: The core model for image and video classification.\n- MTCNN: For face detection in images and videos.\n- OpenCV: For video processing and frame manipulation.\n- MoviePy: For video editing and result generation.\n- Gradio: To create a user-friendly interface for interacting with the deepfake detector.\n\n---\n\n## License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n\n---\n\n## Contributions\n\nContributions are welcome! If you'd like to improve the tool, feel free to submit a pull request or raise an issue.\n\nFor more information, check the [Contribution Guidelines](CONTRIBUTING.md).\n\n---\n\n## References\n- Li et al. (2020): [Celeb-DF(V2)](https://arxiv.org/abs/2008.06456)\n- Rossler et al. (2019): [FaceForensics++](https://arxiv.org/abs/1901.08971)\n- Timesler (2020): [Facial Recognition Model in PyTorch](https://www.kaggle.com/timesler/facial-recognition-model-in-pytorch)\n\n---\n\n### Disclaimer\n\nDeepSecure-AI is a research project and is designed for educational purposes.Please use responsibly and always give proper credit when utilizing the model in your work.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdivith123%2Fdeepsecure-ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdivith123%2Fdeepsecure-ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdivith123%2Fdeepsecure-ai/lists"}