Computer Science PhD student, Queensland University of Technology
I’m a final year PhD student and lecturer at QUT in Brisbane, Australia. My research is focused on deep learning methods for computer vision, for which I mostly use Tensorflow and Python. I’m currently looking at a variety of 3D inference tasks and representations.
When I’m not
cleaning data researching or teaching, I enjoy playing with data visualization tools for the web using dart. I’m currently working on a computer algebra system in dart to help break down complex mathematical processes - think a simpler wolfram alpha’s
show steps button, but free and running client-side.
IGE-Net: Inverse Graphics Energy Networks for Human Pose Estimation and Single-View Reconstruction (CVPR 2019)
Learning Free-Form Deformations for 3D Object Reconstruction (ACCV 2018)
Adversarially Parameterized Optimization for 3D Human Pose Estimation (3DV 2017)
Tensorflow Devpost Competition Entry
- tf_template: supervised learning micro-framework that contains boilerplate for training, evaluating, profiling, and visualizing input/output of models, as well as running basic test.
- tf_toolbox: general testing/profiling functions for tensorflow models.
- tf_marching_cubes: peicewise-differentiable marching cubes implementations.
- tf_nearest_neighbour: brute-force kernels and
tf.py_funchacks for KDTree implementations.
- sdf_renderer: differentiable signed distance function rendering in tensorflow.
- vog_vgg: Models for semantic segmentation based on
VGGnetwork architecture and
- depth_denoising: Models for denoising depth data. Currently supports end-to-end structured prediction energy networks.
Data IO and cleaning is a necessary evil of almost all machine learning research. I maintain the following repositories for loading/manipulating publicly available datasets.
- dids: general interfacing library for saving, loading and lazy manipulation of large datasets
- util3d: common utility functions for manipulating 3D data.
- shapenet: (dataset home page) 3D textured models.
- modelnet: (dataset home page) 3D untextured models.
- seven_scenes: (dataset home page) RGBD / reconstructed TSDF scene dataset.
- nyu: (dataset home page) RGBD semantically segmented scene dataset.
- PASCAL VOC: (dataset home page) RGB semantically segmented scene dataset.
- scannet: (dataset home page) 3D reconstructions of indoor scenes.
- crohme: (dataset home page) Hand written maths expressions.
- human_pose_util: Utility functions for human pose estimation, along with data loading funcitons for Human 3.6m, Human EVA and MPI inf.