Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation

no code implementations • 28 Dec 2022 • Asiri Lindamulage, Nuwan Kodagoda, Shyam Reyal, Pradeepa Samarasinghe, Pratheepan Yogarajah

Magnitude-based pruning is a technique used to optimise deep learning models for edge inference.

Head Pose Estimation regression

2D Pose Estimation based Child Action Recognition

no code implementations • 18 Dec 2022 • Sanka Mohottala, Sandun Abeygunawardana, Pradeepa Samarasinghe, Dharshana Kasthurirathna, Charith Abhayaratne

We present a graph convolutional network with 2D pose estimation for the first time on child action recognition task achieving on par results with an RGB modality based model on a novel benchmark dataset containing unconstrained environment based videos.

2D Pose Estimation Action Recognition +1

Graph Neural Network based Child Activity Recognition

no code implementations • 18 Dec 2022 • Sanka Mohottala, Pradeepa Samarasinghe, Dharshana Kasthurirathna, Charith Abhayaratne

In overcoming the challenges of having small size publicly available child action datasets, several learning methods such as feature extraction, fine-tuning and curriculum learning were implemented to improve the model performance.

Activity Recognition Transfer Learning