arturgontijo/dnn-model-services

Third Party Deep Neural Networks Model Services

A collection of services using third party DNN models.

HTML User's Guide Hub

Getting Started

For more details on how to publish and test a service, select it from the list below:

Audio:

• real-time-voice-cloning (User's Guide) - This service uses Real-Time-Voice-Cloning to clone a voice from a 5 seconds audio file to generate arbitrary speech in real-time. [Reference]
• sound-spleeter (User's Guide) - This service uses Deezer's Spleeter to perform source separation on audio files. [Reference]

   @misc{spleeter2019,
     title={Spleeter: A Fast And State-of-the Art Music Source Separation Tool With Pre-trained Models},
     author={Romain Hennequin and Anis Khlif and Felix Voituret and Manuel Moussallam},
     howpublished={Late-Breaking/Demo ISMIR 2019},
     month={November},
     note={Deezer Research},
     year={2019}
   }
  

Images:

• cntk-image-recon (User's Guide) - This service uses ResNet152 model, trained to recognize different types of flowers and dog breeds. [Reference]
• deepfakes-faceswap (User's Guide) - This service uses the Deepfakes Faceswap, trained on two input videos A and B, to perform face swapping on videos. [Reference]
• deoldify-colorizer (User's Guide) - This service colorizes grayscale images using an AI model. [Reference]
• yolov3-object-detection (User's Guide) - This service uses YOLOv3 model to detect objects on images. [Reference]

    @article{yolov3,
        title={YOLOv3: An Incremental Improvement},
        author={Redmon, Joseph and Farhadi, Ali},
        journal = {arXiv},
        year={2018}
    }
  

• places365-scene-recognition (User's Guide) - This service uses various convolutional neural networks trained on Places365 to perform scene recognition. [Reference]

    @article{zhou2017places,
         title={Places: A 10 million Image Database for Scene Recognition},
         author={Zhou, Bolei and Lapedriza, Agata and Khosla, Aditya and Oliva, Aude and Torralba, Antonio},
         journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
         year={2017},
         publisher={IEEE}
    }
  

• pneumonia-diagnosis (User's Guide) - This service uses VGG19 to classify X-ray chest images. Based on Alishba Imran's work.

Videos:

• i3d-video-action-recognition (User's Guide) - This service uses I3D model to recognize actions on videos (with 400 or 600 labels). [Reference]
• s2vt-video-captioning (User's Guide) - This service uses "Sequence to Sequence - Video to Text" to describe video content with natural language text. [Reference]

    @inproceedings{venugopalan15iccv,
        title = {Sequence to Sequence -- Video to Text},
        author = {Venugopalan, Subhashini and Rohrbach, Marcus and Donahue, Jeff 
                  and Mooney, Raymond and Darrell, Trevor and Saenko, Kate},
        booktitle = {Proceedings of the IEEE International Conference on Computer Vision (ICCV)},
        year = {2015}
    }
  

Games:

• zeta36-chess-alpha-zero (User's Guide) - This service uses AlphaGo Zero methods to learn and play chess. [Reference]

Contributing and Reporting Issues

Please read our guidelines before submitting an issue. If your issue is a bug, please use the bug template pre-populated here. For feature requests and queries you can use this template.

Authors

• Artur Gontijo - Maintainer - SingularityNET
• Ramon Durães - Maintainer - SingularityNET

Licenses

This project is licensed under the MIT License - see the LICENSE file for details.

Each service is licensed as followed:

• cntk-image-recon - MIT License
• deepfakes-faceswap - GPL-3.0
• deoldify-colorizer - MIT License
• i3d-video-action-recognition - Apache License 2.0
• places365-scene-recognition - MIT License
• pneumonia-diagnosis - Attribution 4.0 International (CC BY 4.0)
• real-time-voice-cloning - MIT License
• s2vt-video-captioning - Attribution 4.0 International (CC BY 4.0)
• sound-spleeter - MIT License
• yolov3-object-detection - Public domain
• zeta36-chess-alpha-zero - MIT License