Facial expression identification over a time series of images - QC-202

Preferred Disciplines: Deep Learning, Artificial Intelligence, Computer Vision (Level: Master, PhD or Post-Doc)
Project length: 4-6 months (1 unit)
Approx. start date: As soon as possible
Location: Montreal, QC
No. of Positions: 1
Preferences: None
Company: SeekShift

About Company:

We aim to disrupt the facial motion capture market by leveraging deep neural networks to map, deduce, and identify highly-realistic facial expressions. Our production software will replicate the work of the world’s best animators to yield superior quality facial animations with faster character turnout.

Summary of Project:

  • Identify key facial expressions over a time series of images utilizing CNN and LSTM
  • Interpolation between facial expressions utilizing AI.
  • Head, eye, mouth, and eyelid tracking using computer vision.
  • Sentiment analysis of facial expression (classification)

Research Objectives/Sub-Objectives:

  • Facial expression sentiment analysis and motion interpolation over a time series of images 

Methodology:

    • Identify facial expression over a time series of images
    • Sentiment analysis of facial expression
    • Interpolation between facial expression over time series
    • Output data of facial expression interpolation and sentiment classification over time series.

    Expertise and Skills Needed:

    • Must: Deep Learning (with experience with CNN, RNN, LSTM)
    • Preferable: Artificial Intelligence, Computer Vision

    For more info or to apply to this applied research position, please

    1. Check your eligibility and find more information about open projects.
    2. Interested students need to get the approval from their supervisor and send their CV along with a link to their supervisor’s university webpage by applying through the webform or directly to Marie-Laure de Boutray
    Program: