Benchmarking Air Signature Verifications on T3AAS-v1
Offical PyTorch Implementation of the IJCB 2023 Paper "Enhancing 3D-Air Signature by Pen Tip Tail Trajectory Awareness: Dataset and Featuring by Novel Spatio-temporal CNN" [Arxiv]
Dataset (T3AAS-v1)
Use this Form to request access to the T3AAS-v1 dataset.
Setup
Environment
Use the environment.yaml file to create a conda environment.
Training and Testing Models
Every combination of a model and a dataset needs a separate YAML file similar to config.yaml to run. This file defines most of the specifications regarding the training and testing. The example config.yaml file provided makes all the specifications self-explanatory.
Running one model
Run the run.py file to just run training and testing for one model, whose config file (renamed to config.yaml) is placed in the same directory as the code.
Running multiple models
Create .yaml files for each model and put all such files in a directory called Configs, alongside the directory containing this repository. Run the autorun.py file to sequentially run the training and testing for each model.
About Us
Visit the following site - Machine Intelligence Group.