Applied Research Scientist, Multimodal Video Understanding (Deep Fake Detection) at Facebook
New York City, NY, US
Every day, massive amounts of video are uploaded into Facebook's services. In order to serve our communities better, it is critical that we can identify maliciously tampered videos, for example videos where politicians are made to say things they never actually said in real life, or videos that show events which never occurred. The ability to automatically spot such material will help keep elections fair and help prevent the malicious manipulation of our communities. This position is in the Multimodal Video Understanding Team, which is part of Facebook AI. The team carries out applied research in ML/AI and designs, develops and deploys state of the art ML/AI algorithms to the rest of Facebook. In addition to tamper-detection, our algorithms are used for ranking, recommendations, advertising, content filtering, and power multiple product experiences across Facebook, Messenger, Instagram, WhatsApp and Oculus.
RESPONSIBILITIES
  •  
     
    Use multi-modal signals to develop methods for the detection of tampered video content
  •  
     
    Develop highly scalable algorithms based on state-of-the-art machine learning and neural network methodologies
  •  
     
    Conduct research to advance the state-of-the-art, and publish work in relevant speech, vision, NLP, and machine learning conferences and journals
  •  
     
    Apply expert coding to projects in partnership with other engineers across research, product, and infrastructure teams
  •  
     
    Adapt machine learning and neural network algorithms for training competitive, state-of-the-art models which make the best use of modern parallel environments (e.g. distributed clusters, GPU)
MINIMUM QUALIFICATIONS
  •  
     
    PhD in Computer Science or related quantitative field
  •  
     
    Experience with machine learning, neural networks, and deep learning
  •  
     
    Experience building systems based on machine learning and/or deep learning methods, especially in the areas of speech recognition, natural language processing, image processing, or other machine-perception tasks
  •  
     
    Experience developing and debugging in C/C++ and/or Python
Facebook is committed to providing reasonable accommodations for qualified individuals with disabilities and disabled veterans in our job application procedures. If you need assistance or an accommodation due to a disability, you may contact us at accommodations-ext@fb.com.