Our aim is to deeply investigate novel computer vision and artificialintelligence (AI) to identify human emotions for advancing the emotionalintelligence of AI-based solutions and accelerate innovations for real-worldapplications. We research expressed and suppressed visual cues, unseenphysiological signals and privacy preserving methods.


Key results:

  • Developed 3D body micro-gesture mesh datasets and online recognition methods that facilitates 3D modelling of human micro-gestures for hidden emotion analysis.
  • Developed 4D facial micro-expression dataset with 22 action units (AUs) and 5 categories of emotion labels with baseline results to facilitate micro-expession recognition.
  • Proposed several new machine learning methods and network architectures with excellent performances on challenging 3D synthesizing task.
  • Explored new methodology for privacy protection.


Computational resources:

  • Size of training set: 5-10 datasets, 20G-150G per dataset

  • Computing resources: ~0.6M BUs on Puhti, ~1.5M BUs on Mahti, 2.3M BUs in total
  • Storage resources: ~1TB on Puhti, ~200G on Allas, ~1TB on Mahti
  • Model training time (hours/days/months): 10-20 different models, 10 hours-72 hours per model
  • Any other significant resource details: 1. Implementing the training on LUMI supercomputer cluster

                                                                          2. Hosting the project websites





Selected Publications:

  1. H Mo, G Zhao. RIC-CNN: rotation-invariant coordinate convolutional neural network. Pattern Recognition. 2024

  2. H Chen, H Tang, R Timofte, L Van Gool, G Zhao. LART: Neural Correspondence Learning with Latent Regularization Transformer for 3D Motion Transfer,NeurIPS 2023

  3. Y Liu, X Zhang, J Kauttonen, G Zhao, Uncertain Facial Expression Recognition via Multi-task Assisted Correction, IEEE Transactions on Multimedia 2023

  4. H Chen, H Shi, X Liu, X Li, G Zhao, SMG: A Micro-Gesture Dataset Towards Spontaneous Body Gestures for Emotional Stress State Analysis, International Journal of Computer Vision 2023

  5. T Varanka, Y Li, W Peng, G Zhao, Data Leakage and Evaluation Issues in Micro-Expression Analysis, IEEE Transactions on Affective Computing, 2023

  6. M Savic, G Zhao. De-identification of facial videos while preserving remote physiological utility. Proceedings of the British Machine Vision Conference (BMVC), 2023​

  7. G Zhao, Y Li , Q Xu. From Emotion AI to Cognitive AI. International Journal of Network Dynamics and Intelligence (IJNDI)​

  8. K H M Cheng, Z Yu, H Chen, G Zhao. Benchmarking 3D Face De-identification with Preserving Facial Attributes. IEEE International Conference on Image Processing, ICIP 2022​

  9. Y Li, J Wei, Y Liu, J Kauttonen, G Zhao. Deep Learning for Micro-expression Recognition: A Survey. IEEE Transactions on Affective Computing, Vol. 13, NO. 4​

  10. H Chen, H Tang, Z Yu,N Sebe, G Zhao. Geometry-Contrastive Transformer for Generalized 3D Pose Transfer. Proceedings of the AAAI Conference on ArtificialIntelligence (AAAI), 2022​

  11. H Chen, H Tang, N Sebe, G Zhao. AniFormer: Data-driven 3D Animation with Transformer. Proceedings of the British Machine Vision Conference (BMVC), 2021











         

  • No labels

Korkeakoulutuksen ja tutkimuksen yhteistyöalustan tietosuojaseloste