See what we're up to
Seeing the Pose in the Pixels: Learning Pose-Aware Representations in Vision Transformers
Two approaches for training ViTs to learn Pose-aware representations for ADL videos, enabling fine-grained and viewpoint-agnostic visual perception.
Few-Shot Keyword Spotting
Metric learning algorithm and few-shot Google Speech Command dataset to enable few-shot keyword spotting.
Error-related Potential (ErrP) Variability
Classification and statistical analysis on variability and transferability of error-related EEG brain signal patterns, ErrP.
Deep Reinforcement Learning (DRL) Monitor
A monitoring tool of deep reinforcement learning algorithms to analyze its learning process and provide example-based explanation